JP7591192B2 - Inspection system, inspection method, and method for manufacturing an inspection system - Google Patents

Description

Embodiments of the present disclosure relate to an inspection system, an inspection method, and a method for manufacturing an inspection system.

Systems and methods for non-destructively inspecting objects have been proposed. For example, Patent Document 1 discloses a method for inspecting the composition of resin based on the reflection spectrum obtained by irradiating light onto the resin.

JP 2019-86412 A

The spectrum may change depending on the ambient temperature.

The embodiment of the present disclosure aims to provide an inspection system, an inspection method, and a method for manufacturing an inspection system that can determine the composition of an object while taking into account the effects of ambient temperature.

One embodiment of the present disclosure is an inspection system for determining a composition of a resin layer of a resin sheet, comprising:
a temperature monitor for monitoring an ambient temperature;
An irradiation device including an irradiation unit that irradiates the resin sheet with near-infrared light as irradiation light;
a detection device that detects the intensity of light reflected by the resin sheet or light transmitted through the resin sheet;
a generation device for generating a spectrum containing information about the intensities at a plurality of wavelength points;
a processing device that calculates, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
The inspection system includes a determination device including: a memory unit in which a standard regression equation expressing a relationship between a plurality of parameter values and a first candidate composition and a second candidate composition; an analysis unit that applies the standard regression equation to the plurality of parameter values to calculate a similarity; and a determination unit that determines whether the composition of the resin layer is the first candidate composition or the second candidate composition based on a comparison between the similarity and a threshold corresponding to the ambient temperature.

In an inspection system according to an embodiment of the present disclosure, the second candidate composition corresponds to a 21st candidate composition and a 22nd candidate composition, the storage unit stores a sub-standard regression equation that represents a relationship between a parameter value at a standard temperature and the 21st candidate composition and the 22nd candidate composition, and when the composition of the resin layer is determined to be the second candidate composition, the analysis unit applies the sub-standard regression equation to the multiple parameter values to calculate a sub-similarity, and the determination unit may determine whether the composition of the resin layer is the 21st candidate composition or the 22nd candidate composition based on a comparison between the sub-similarity and a sub-threshold corresponding to the ambient temperature.

One embodiment of the present disclosure is an inspection system for determining a composition of a resin layer of a resin sheet, comprising:
a temperature monitor for monitoring an ambient temperature;
An irradiation device including an irradiation unit that irradiates the resin sheet with near-infrared light as irradiation light;
a detection device that detects the intensity of light reflected by the resin sheet or light transmitted through the resin sheet;
a generation device for generating a spectrum containing information about the intensities at a plurality of wavelength points;
a processing device that calculates, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
The inspection system includes a determination device including: a memory unit in which a plurality of regression equations are stored that express relationships between a parameter value and a first candidate composition and a second candidate composition at a plurality of temperatures; an analysis unit that selects a regression equation corresponding to the ambient temperature from the plurality of regression equations and applies the selected regression equation to the plurality of parameter values to calculate a similarity; and a determination unit that determines whether a composition of the resin layer is the first candidate composition or the second candidate composition based on the similarity.

In an inspection system according to an embodiment of the present disclosure, the second candidate composition corresponds to a 21st candidate composition and a 22nd candidate composition, the storage unit stores a plurality of sub-regression equations that represent the relationship between a parameter value and the 21st candidate composition and the 22nd candidate composition at a plurality of temperatures, and when the composition of the resin layer is determined to be the second candidate composition, the analysis unit selects a sub-regression equation that corresponds to the ambient temperature from the plurality of sub-regression equations and applies the selected sub-regression equation to the plurality of parameter values to calculate a sub-similarity, and the determination unit may determine whether the composition of the resin layer is the 21st candidate composition or the 22nd candidate composition based on the sub-similarity.

An inspection system according to one embodiment of the present disclosure may include a cooling unit that cools the irradiation device and/or the detection device.

One embodiment of the present disclosure is an inspection system for determining a composition of a resin layer of a resin sheet, comprising:
An irradiation device including an irradiation unit that irradiates the resin sheet with near-infrared light as irradiation light;
a detection device that detects the intensity of light reflected by the resin sheet or light transmitted through the resin sheet;
A cooling unit that cools the irradiation device and/or the detection device;
a generation device for generating a spectrum containing information about the intensities at a plurality of wavelength points;
a processing device that calculates, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
The inspection system includes a determination device including: a memory unit in which a regression equation expressing a relationship between a parameter value and a first candidate composition and a second candidate composition is stored; an analysis unit that applies the regression equation to the multiple parameter values to calculate a similarity; and a determination unit that determines whether the composition of the resin layer is the first candidate composition or the second candidate composition based on the similarity.

In an inspection system according to one embodiment of the present disclosure, the processing device may include 69 or more parameter values of the resin layer within a wavelength range of 800 nm or more and 2500 nm or less.

In an inspection system according to an embodiment of the present disclosure, the irradiation device may further include a diffusion section that diffuses the irradiation light.

In an inspection system according to one embodiment of the present disclosure, the detection device may include a near-infrared spectrometer.

In an inspection system according to one embodiment of the present disclosure, the detection device may include a near-infrared hyperspectral camera.

One embodiment of the present disclosure is an inspection method for determining a composition of a resin layer of a resin sheet, comprising:
A first irradiation step of irradiating the resin sheet with near-infrared light as irradiation light;
a first detection step of detecting an intensity of light reflected by the resin sheet or light transmitted through the resin sheet;
a first generation step of generating a spectrum including information about the intensities at a plurality of wavelength points;
a first processing step of calculating, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained at a plurality of wavelength points by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
an analysis step of calculating a similarity by applying a standard regression equation that represents a relationship between a parameter value at a standard temperature and a first candidate composition and a second candidate composition to the plurality of parameter values;
and a determination step of determining whether the composition of the resin layer is the first candidate composition or the second candidate composition based on a comparison of the similarity with a threshold corresponding to the ambient temperature.

In an inspection method according to an embodiment of the present disclosure, the second candidate composition corresponds to the 21st candidate composition and the 22nd candidate composition, and when the composition of the resin layer is determined to be the second candidate composition, the analysis step includes a step of calculating sub-similarity by applying a sub-standard regression equation representing the relationship between the parameter value at standard temperature and the 21st candidate composition and the 22nd candidate composition to the multiple parameter values, and the determination step may include a step of determining whether the composition of the resin layer is the 21st candidate composition or the 22nd candidate composition based on a comparison between the sub-similarity and a sub-threshold corresponding to the ambient temperature.

One embodiment of the present disclosure is an inspection method for determining a composition of a resin layer of a resin sheet, comprising:
A first irradiation step of irradiating the resin sheet with near-infrared light as irradiation light;
a first detection step of detecting an intensity of light reflected by the resin sheet or light transmitted through the object;
a first generation step of generating a spectrum including information about the intensities at a plurality of wavelength points;
a first processing step of calculating, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained at a plurality of wavelength points by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
an analysis step of selecting a regression equation corresponding to an ambient temperature from among a plurality of regression equations that express relationships between a parameter value and a first candidate composition and a second candidate composition at a plurality of temperatures, and applying the selected regression equation to the plurality of parameter values to calculate a similarity;
and a determining step of determining whether the composition of the resin layer is the first candidate composition or the second candidate composition based on the similarity.

In an inspection method according to an embodiment of the present disclosure, the second candidate composition corresponds to the 21st candidate composition and the 22nd candidate composition, and when the composition of the resin layer is determined to be the second candidate composition, the analysis step includes a step of selecting a sub-regression equation corresponding to the ambient temperature from among a plurality of sub-regression equations that represent the relationship between a parameter value and the 21st candidate composition and the 22nd candidate composition at a plurality of temperatures, and applying the selected sub-regression equation to the plurality of parameter values to calculate sub-similarity, and the determination step may include a step of determining whether the composition of the resin layer is the 21st candidate composition or the 22nd candidate composition based on the sub-similarity.

In an inspection method according to one embodiment of the present disclosure, the first processing step may calculate 69 or more parameter values of the resin layer within a wavelength range of 800 nm or more and 2500 nm or less.

In an inspection method according to one embodiment of the present disclosure, the first irradiation step may include diffusing the irradiation light.

In an inspection method according to one embodiment of the present disclosure, the first detection step may involve acquiring the spectrum using a near-infrared spectrometer.

In an inspection method according to one embodiment of the present disclosure, the first detection step may involve acquiring the spectrum using a near-infrared hyperspectral camera.

One embodiment of the present disclosure is a method for manufacturing an inspection system, comprising:
a standard sample irradiating step of irradiating a first standard sample including a first candidate composition and a second standard sample including a second candidate composition with near-infrared light as irradiation light at a plurality of temperatures including a standard temperature;
a standard sample detection step of receiving light reflected by the first standard sample and the second standard sample or light transmitted through the first standard sample and the second standard sample at a plurality of temperatures, respectively, and acquiring a first standard spectrum and a second standard spectrum, each of which includes information on the intensity of the light at a plurality of wavelength points;
a standard sample processing step of calculating values of the light intensity at a plurality of wavelength points of the standard spectrum, or values obtained at a plurality of wavelength points by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on each of the standard spectra, as a plurality of first standard parameter values and a plurality of second standard parameter values corresponding to each wavelength point, for each temperature;
a regression equation creating step of creating a standard regression equation that expresses a relationship that holds at the standard temperature between parameter values at a plurality of wavelength points and a first candidate composition and a second candidate composition by analyzing the plurality of first standard parameter values and the plurality of second standard parameter values at the standard temperature using a multivariate analysis technique;
and a threshold creating process for calculating, for each temperature, a threshold for distinguishing between the first candidate composition and the second candidate composition based on a similarity calculated by applying the standard regression equation to the plurality of first standard parameter values and the plurality of second standard parameter values at a temperature other than the standard temperature.

One embodiment of the present disclosure is a method for manufacturing an inspection system, comprising:
a standard sample irradiating step of irradiating a first standard sample including a first candidate composition and a second standard sample including a second candidate composition with near-infrared light as irradiation light at a plurality of temperatures including a standard temperature;
a standard sample detection step of receiving light reflected by the first standard sample and the second standard sample or light transmitted through the first standard sample and the second standard sample at a plurality of temperatures, respectively, and acquiring a first standard spectrum and a second standard spectrum, each of which includes information on the intensity of the light at a plurality of wavelength points;
a standard sample processing step of calculating values of the light intensity at a plurality of wavelength points of the standard spectrum, or values obtained at a plurality of wavelength points by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on each of the standard spectra, as a plurality of first standard parameter values and a plurality of second standard parameter values corresponding to each wavelength point, for each temperature;
and a regression equation creation process for creating, for each temperature, a regression equation that represents a relationship between parameter values at multiple wavelength points and a first candidate composition and a second candidate composition by analyzing the multiple first standard parameter values and the multiple second standard parameter values for each temperature using a multivariate analysis technique.

One embodiment of the present disclosure is a method for manufacturing an inspection system, comprising:
a standard sample irradiating step of irradiating a first standard sample including a first candidate composition and a second standard sample including a second candidate composition with near-infrared light as irradiation light at a plurality of temperatures;
a standard sample detection step of receiving light reflected by the first standard sample and the second standard sample or light transmitted through the first standard sample and the second standard sample at a plurality of temperatures, respectively, and acquiring a first standard spectrum and a second standard spectrum, each of which includes information on the intensity of the light at a plurality of wavelength points;
a standard sample processing step of calculating values of the light intensity at a plurality of wavelength points of the standard spectrum, or values obtained at a plurality of wavelength points by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on each of the standard spectra, as a plurality of first standard parameter values and a plurality of second standard parameter values corresponding to each wavelength point, for each temperature;
and a regression equation creation process for creating a general-purpose regression equation that expresses a relationship between parameter values at multiple wavelength points and a first candidate composition and a second candidate composition, the relationship being valid at any of multiple temperatures, by analyzing the multiple first standard parameter values and the multiple second standard parameter values for each temperature using a multivariate analysis technique.

According to one embodiment of the present disclosure, the composition of an object can be determined taking into account the effect of ambient temperature.

FIG. 2 is a cross-sectional view showing a resin sheet according to one embodiment. FIG. 1 is a block diagram illustrating an inspection system according to an embodiment. FIG. 2 is a schematic diagram illustrating an illumination device and a detection device of an inspection system according to an embodiment. FIG. 2 is a perspective view showing an example of a case that houses an irradiation device and a detection device. FIG. 1 illustrates an example of how the inspection system can be used. FIG. 1 is a diagram showing an example of a standard reflectance spectrum. FIG. 13 is a diagram showing the results of analysis using a multivariate analysis method. FIG. 2 shows the results of applying a standard regression equation to parameters of a first standard and a second standard at a first temperature. FIG. 13 shows the results of applying a standard regression equation to parameters of a first standard and a second standard at a second temperature. FIG. 13 shows the results of applying a standard regression equation to parameters of a first standard and a second standard at a third temperature. FIG. 13 shows the results of applying a standard regression equation to parameters of a first standard sample and a second standard sample at a fourth temperature. FIG. 11 is a diagram showing calculation results of threshold values at first to fourth temperatures. FIG. 13 is a diagram showing the results of determining the composition of a sample using a standard regression equation and first to fourth threshold values. FIG. 13 is a diagram showing an example of a calibration spectrum. FIG. 13 is a diagram showing an example of a process for irradiating a calibration sample with light. FIG. 1 is a block diagram illustrating an inspection system according to an embodiment. FIG. 1 is a block diagram illustrating an inspection system according to an embodiment. FIG. 13 is a diagram showing the calculation results of a general regression equation. FIG. 13 is a diagram showing the results of determining the composition of a sample using a universal regression equation and a universal threshold value. 10 is a flowchart showing an example of an inspection process. FIG. 4 is a diagram showing an example of a target reflectance spectrum. FIG. 13 is a diagram illustrating an example of an averaged spectrum. FIG. 13 is a diagram showing an example of the difference between a target reflectance spectrum and an averaged spectrum. FIG. 13 is a diagram showing a measurement result of a first object. FIG. 11 is a diagram showing the measurement results of a second object. FIG. 13 is a diagram showing the measurement results of a third object. FIG. 13 is a diagram showing the measurement results of a fourth object. FIG. 13 is a diagram showing the measurement results of a fifth object. FIG. 13 is a diagram showing the measurement results of a sixth object. FIG. 13 is a diagram showing an example of the results of determining the composition of a sample. FIG. 1 is a block diagram illustrating an inspection system according to an embodiment. 13 is a flowchart illustrating an example of a calibration process. FIG. 13 is a diagram showing the measurement results of a first calibration sample. FIG. 13 is a diagram showing the measurement results of a first calibration sample. FIG. 13 is a diagram showing the measurement results of a second calibration sample. FIG. 13 is a diagram showing the measurement results of a second calibration sample. FIG. 13 is a diagram showing an example of the results of determining the composition of a sample. FIG. 13 is a diagram showing an example of the results of determining the composition of a sample. FIG. 13 is a diagram showing an example of the results of determining the composition of a sample. FIG. 2 is a perspective view showing an example of a case. FIG. 2 is a perspective view showing an example of a case. FIG. 2 is a perspective view showing an example of a case.

The configuration of an object and an inspection system according to an embodiment of the present disclosure will be described in detail with reference to the drawings. Note that the embodiment described below is an example of an embodiment of the present disclosure, and the present disclosure should not be interpreted as being limited to these embodiments. In this specification, terms such as "substrate" and "sheet" are not distinguished from each other based only on the difference in name. For example, "substrate" is a concept that includes members that may be called sheets or films.

Terms used in this specification that specify shapes, geometric conditions, and their degrees, such as "parallel" and "orthogonal," as well as values of lengths and angles, are not limited to their strict meanings, but are interpreted to include the range within which similar functions can be expected.

In this specification, when multiple upper limit candidates and multiple lower limit candidates are given for a certain parameter, the numerical range of the parameter may be constructed by combining any one of the upper limit candidates and any one of the lower limit candidates. For example, consider a case where "Parameter B is, for example, A1 or more, may be A2 or more, or may be A3 or more. Parameter B is, for example, A4 or less, may be A5 or less, or may be A6 or less." In this case, the numerical range of parameter B may be A1 or more and A4 or less, A1 or more and A5 or less, A1 or more and A6 or less, A2 or more and A4 or less, A2 or more and A5 or less, A2 or more and A6 or less, A3 or more and A4 or less, A3 or more and A5 or less, or A3 or more and A6 or less.

In the drawings referred to in this embodiment, identical parts or parts having similar functions are given the same or similar symbols, and repeated explanations may be omitted. In addition, the dimensional ratios of the drawings may differ from the actual ratios for the convenience of explanation, and some components may be omitted from the drawings.

(First embodiment)
1 is a diagram showing an example of an object 10. The object 10 is, for example, a resin sheet. Resin sheets are used for various purposes, such as decorative sheets for building materials such as flooring materials, agricultural sheets, encapsulation sheets for semiconductor elements, and packaging sheets.

The object 10 includes a base sheet 11 and a resin layer 12 laminated on the base sheet 11. The resin layer 12 may be provided over the entire surface of the base sheet 11, or may be provided partially on the surface of the base sheet 11 so as to present a pattern. In the following description, the surface of the object 10 located on the resin layer 12 side is referred to as the first surface 10x, and the surface of the object 10 located on the base sheet 11 side is referred to as the second surface 10y.

The base sheet 11 includes, for example, paper, synthetic resin, etc. More specifically, the base sheet 11 may include a cellulose resin, or an olefin-based synthetic resin such as polypropylene. The basis weight of the base sheet 11 is, for example, 50 g/m ² or more. The basis weight of the base sheet 11 is, for example, 300 g/m ² or less, and may be 120 g/m ² or less.

The resin layer 12 includes a resin material. The resin material included in the resin layer 12 is not particularly limited, but is, for example, a polymer resin. Examples of the polymer resin include polyethylene resin, polyvinyl chloride resin (PVC), polyvinyl alcohol resin (PVA), polypropylene, polystyrene, polyvinyl acetate, acrylic resin, polyethylene terephthalate, polyester, polyamide, polycarbonate, polyurethane, and polyimide.
Examples of polyethylene-based resins include polyethylene, as well as ethylene copolymers having ethylene and a component other than ethylene as monomers, such as ethylene-vinyl acetate copolymer resin (EVA) and ethylene-α-olefin copolymer. The resin layer 12 may also contain rubber. Examples of rubber include isoprene rubber, butadiene rubber, acrylonitrile-butadiene copolymer rubber, and styrene-butadiene copolymer rubber. The thickness of the resin layer 12 is, for example, 40 μm or more. The thickness of the resin layer 12 may be 700 μm or less.
As described later, the resin layer 12 may contain a foaming agent or a foaming assistant in addition to the resin material. In this case, the thickness of the resin layer 12 before foaming is, for example, 40 μm or more and 100 μm or less. The thickness of the resin layer 12 after foaming is, for example, 300 μm or more and 700 μm or less.

The resin layer 12 may further contain a coloring material such as a pigment. The pigment may be an inorganic pigment or an organic pigment. Examples of inorganic pigments include titanium oxide, zinc oxide, carbon black, black iron oxide, yellow iron oxide, yellow lead, molybdate orange, cadmium yellow, nickel titanium yellow, chrome titanium yellow, iron oxide (red oxide), cadmium red, ultramarine, Prussian blue, cobalt blue, chromium oxide, cobalt green, aluminum powder, bronze powder, titanium mica, and zinc sulfide. Examples of organic pigments include aniline black, perylene black, azo-based (azo lake, insoluble azo, condensed azo), polycyclic (isoindolinone, isoindoline, quinophthalone, perinone, flavanthrone, anthrapyrimidine, anthraquinone, quinacridone, perylene, diketopyrrolopyrrole, dibromoanzanthrone, dioxazine, thioindigo, phthalocyanine, indanthrone, and halogenated phthalocyanine. The pigment content in the resin layer 12 may be, for example, 5 parts by mass or more, or 15 parts by mass or more, relative to 100 parts by mass of the resin component. The pigment content in the resin layer 12 may be, for example, 50 parts by mass or less, or 30 parts by mass or less, relative to 100 parts by mass of the resin component.

The resin layer 12 may contain a filler. This can increase the strength, hardness, etc. of the resin layer 12. The filler includes inorganic materials such as calcium carbonate, aluminum hydroxide, magnesium hydroxide, antimony trioxide, zinc borate, molybdenum compounds, and talc. The amount of filler contained in the resin layer 12 may be, for example, 0.1 parts by mass or more, or 20 parts by mass or more, relative to 100 parts by mass of the resin material. The amount of filler contained in the resin layer 12 may be, for example, 100 parts by mass or less, or 70 parts by mass or less, relative to 100 parts by mass of the resin material.

The resin layer 12 may contain additives such as antifungal agents, insect repellents, preservatives, antibacterial agents, deodorants, polymerization initiators, polymerization inhibitors, sensitizers, crosslinking agents, plasticizers, flame retardants, charge control agents, heat stabilizers, light stabilizers, conductive agents, defoamers, rust inhibitors, antioxidants, foaming agents, foaming assistants, near-infrared absorbers, ultraviolet absorbers, and emulsifiers.

The object 10 may be an item other than a resin sheet. For example, the object 10 may include paper, metal, ceramics, medicine, etc. That is, the object to be inspected by the inspection system 15 and inspection method described below may be paper, metal, ceramics, medicine, etc.

In this specification, the object 10 is a standard specimen or a sample. A standard specimen includes a known candidate composition. A sample includes an unknown candidate composition. An inspection system 15, which will be described later, determines the candidate composition contained in the sample based on information obtained by inspecting the standard specimen. The inspection system 15 will be described below.

[Inspection system]
2 is a block diagram showing an inspection system 15 for inspecting the object 10. The inspection system 15 measures a spectrum that represents a characteristic of the resin layer 12. The inspection system 15 can determine which of a plurality of candidate compositions the composition of the material constituting the resin layer 12 is based on the spectrum. For example, the inspection system 15 can determine whether the composition of the material constituting the resin layer 12 is a first candidate composition or a second candidate composition. Information on the plurality of candidate compositions is stored in advance in the inspection system 15.

The candidate composition may represent the resin material contained in the resin layer 12 as a main component, or may represent the resin material contained in the resin layer 12 as a secondary component. The main component is the resin material in the resin layer 12 that has the highest content. The secondary component is the resin material in the resin layer 12 that has a lower content than the main component. The candidate composition may represent the type of additive added to the resin layer 12, or may represent whether or not the resin layer 12 contains an additive.

The inspection system 15 includes an irradiation device 21, a temperature monitor 23, a detection device 24, a generation device 25, a processing device 27, and a determination device 28. The irradiation device 21 and the detection device 24 constitute a spectroscopic module 20, which will be described later. The spectroscopic module 20 is housed in a case 40, which will be described later. Some or all of the generation device 25, the processing device 27, and the determination device 28 may be included in the spectroscopic module 20, or may be included in a component other than the spectroscopic module 20. For example, some or all of the generation device 25, the processing device 27, and the determination device 28 may be realized by a computer capable of communicating with the spectroscopic module 20.

[Irradiation device]
The irradiation device 21 irradiates the object 10 with light such as near-infrared light. The light irradiated to the object 10 is also referred to as irradiated light. The irradiation device 21 includes an irradiation unit 211. The irradiation unit 211 irradiates the first surface 10x of the object 10 with light. The wavelength of the light is, for example, 800 nm or more, and may be 900 nm or more. The wavelength of the light is, for example, 2500 nm or less, may be 2000 nm or less, or may be 1700 nm or less.

FIG. 3 is a diagram showing a schematic of the irradiation device 21 and the detection device 24. The irradiation device 21 may include a diffusion section 212. The diffusion section 212 diffuses the light emitted from the irradiation section 211. This allows the light L1 to be irradiated onto the object 10 from various directions. This allows the inspection system 15 to obtain a diffuse reflectance spectrum. The diffusion section 212 is, for example, a diffusion plate disposed between the irradiation section 211 and the object 10.

[Detection device]
The detection device 24 receives the light L2 reflected by the object 10. The detection device 24 can detect the intensity of the light L2. In the following description, the light reflected by the object 10 is also referred to as reflected light L2.

The detection device 24 includes a detection unit 241 that detects reflected light L2 at multiple wavelength points. The number of wavelength points is determined according to the resolution of the detection unit 241. For example, if the detection unit 241 detects the intensity of reflected light L2 in the range of 900 nm to 1800 nm with a resolution of 3 nm, the number of wavelength points is 301. A near-infrared spectrometer, a near-infrared hyperspectral camera, or the like can be used as the detection unit 241.

The detection unit 241 detects the reflected light L2 specularly reflected by the object 10. The detection unit 241 may detect the reflected light L2 diffusely reflected by the object 10 in addition to the reflected light L2 specularly reflected by the object 10. This makes it possible to obtain more information about the resin material. This improves the accuracy of determining the composition of the resin layer 12. In FIG. 3, the symbol H1 represents the distance between the detection unit 241 and the object 10, and the symbol W1 represents the width of the detection unit 241. The distance H1 and the width W1 are appropriately determined so that the reflected light L2 diffusely reflected by the object 10 is incident on the detection unit 241.

[Temperature monitor]
The temperature monitor 23 monitors the ambient temperature of the inspection system 15. The temperature monitor 23 may be included in the spectroscopic module 20. For example, the spectroscopic module 20 may include a substrate and the temperature monitor 23 mounted on the substrate. In this case, the temperature monitor 23 is, for example, a temperature sensor IC. The temperature sensor IC monitors the temperature inside the spectroscopic module 20. The temperature monitor 23 may be housed in a case 40 described later. The temperature monitor 23 may have a structure separated from the spectroscopic module 20 and the case 40. For example, a thermometer installed in a room in which the case 40 is placed may be used as the temperature monitor 23. The temperature monitor 23 transmits information on the ambient temperature to the determination device 28.

The determination device 28 may obtain information from two or more temperature monitors 23. For example, the determination device 28 may obtain information from a temperature monitor 23 located inside the spectroscopic module 20 and information from a temperature monitor 23 located outside the case 40.
The determination device 28 may use one of two or more pieces of information as the ambient temperature. For example, when the change in the spectrum is mainly caused by a temperature change in the spectroscopic module 20, the determination device 28 may use the information from the temperature monitor 23 located inside the spectroscopic module 20 as the ambient temperature. For example, when the change in the spectrum is mainly caused by a temperature change in the object 10, the determination device 28 may use the information from the temperature monitor 23 located outside the case 40 as the ambient temperature.
The determination device 28 may use the result of processing two or more pieces of information as the ambient temperature. For example, the determination device 28 may use the average value of the measurement result by the temperature monitor 23 located inside the spectroscopic module 20 and the measurement result by the temperature monitor 23 located outside the case 40 as the ambient temperature.

[Generation device]
The generating device 25 includes a first generating unit 251 and a second generating unit 252. The first generating unit 251 generates a spectrum including information about the intensity of the reflected light L2 at a plurality of wavelength points. In the following description, the spectrum including information about the intensity of the reflected light L2 reflected by the object 10 is also referred to as the object reflection spectrum. The object reflection spectrum when the object 10 is a standard sample is also referred to as the standard reflection spectrum. The object reflection spectrum when the object is a sample is also referred to as the sample reflection spectrum.

In this specification, "spectrum" means a group of data corresponding to a wavelength point. When the group of data is illustrated, a graph such as that shown in FIG. 6 described later is generated. The data includes information regarding the intensity of reflected light L2 at the wavelength point. The data may be the intensity of reflected light L2 itself. The data may be calculated by performing some processing on the intensity of reflected light L2 at the wavelength point. For example, the data may be absorbance. Absorbance is a dimensionless quantity that represents the degree to which the object 10 absorbs light. The absorbance Abs is calculated, for example, by the following formula.
Abs=-log ₁₀ (Itar/Iref)
Itar is the intensity of light reflected by the object 10. Iref is the intensity of light reflected by a calibration sample. The calibration sample is, for example, a standard reflector. The standard reflector is also called a standard white plate or a white plate.

The second generation unit 252 generates a spectrum containing information about the intensity of light reflected by the calibration sample at multiple wavelength points. In the following description, the spectrum containing information about the intensity of light reflected by the calibration sample is also referred to as a calibration spectrum. The first generation unit 251 may generate the target reflection spectrum by using the calibration spectrum as a reference, as expressed by the above formula.

[Processing Device]
The processing device 27 includes a first processing unit 271. The first processing unit 271 performs a predetermined process on the target reflection spectrum generated by the generating device 25. In the following description, values obtained at multiple wavelength points by performing a predetermined process on the reflection spectrum are referred to as parameter values. Note that the concept of "performing a process on the target reflection spectrum" includes not only the case where the target reflection spectrum itself is processed, but also the case where a spectrum containing information equivalent to the target reflection spectrum, such as the above-mentioned absorbance spectrum, is processed.

The processing device 27 performs processing on the target reflectance spectrum, including, for example, averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof. This allows multiple parameter values corresponding to multiple wavelength points to be calculated.

The averaging process will be described. First, the irradiation device 21 irradiates irradiation light onto multiple regions on the first surface 10x side of the object 10 whose state is to be determined. This allows the object reflection spectrum to be obtained in each region. Next, the processing device 27 adds up the values of the multiple object reflection spectra for each wavelength point. Next, the added value is divided by the number of object reflection spectra. This allows the parameter value to be calculated after the averaging process. The number of regions to which the irradiation light is irradiated is not particularly limited, but is, for example, three or more. By performing the averaging process, the composition of the resin layer 12 can be determined more accurately even if the surface of the object 10 is uneven or the surface condition is not uniform.

Smoothing is a process in which, when the value of the target reflectance spectrum at a certain wavelength point deviates from the value of the target reflectance spectrum at other wavelength points, the deviating value is removed or brought closer to the other value. This makes it possible to obtain a target reflectance spectrum with fewer overall peaks.

Scattering correction is a process in which, when measuring the target reflectance spectrum of a scattering object, the scattering factor is estimated using the least squares method to obtain the optimal target reflectance spectrum.

Baseline correction is the process of subtracting the baseline from the acquired target reflectance spectrum.

Peak shift correction is a process that corrects the shift in the position of the peak in the acquired target reflectance spectrum.

The processing by the processing device 27 may have the following purpose. The object reflection spectrum acquired by the detection device contains information on the absolute value of the intensity of the reflected light, and therefore varies depending on the output of the irradiation unit 211, the distance between the irradiation unit 211 and the object 10, the distance between the object 10 and the detection unit 241, and the like. As a result, the determination of the composition of the resin layer 12 may vary. The processing by the processing device 27 may aim to reduce such variation. The process for reducing variation is not particularly limited as long as it is a process that can remove or reduce information on the absolute value of the intensity of the reflected light, and is, for example, differentiation or normalization. When differentiation is performed as a process for reducing variation, the number of differentiations is not particularly limited, and for example, second-order differentiation can be performed.

The number of parameter values possessed by the processing device 27 and the wavelength range in which the parameter values are determined are not particularly limited, so long as the composition of the resin layer 12 can be determined accurately based on the parameter values. For example, the processing device preferably includes 69 or more parameter values in a wavelength range of 800 nm or more and 2500 nm or less. Similarly, the distance between two adjacent wavelength points is not particularly limited, so long as the composition of the resin layer 12 can be determined accurately based on the parameter values. The distance between two adjacent wavelength points is, for example, 20 nm or less, may be 10 nm or less, or may be 5 nm or less.

(Determination device)
The determination device 28 determines the composition of the resin layer 12 based on the parameter values calculated by the processing device 27. Specifically, the determination device 28 determines which of a plurality of candidate compositions the composition of the material constituting the resin layer 12 of the object 10 is, based on the parameter values calculated by the processing device 27. In the present embodiment, the determination device 28 determines whether the composition of the material constituting the resin layer 12 is the first candidate composition or the second candidate composition. As shown in FIG. 2, the determination device 28 includes a storage unit 281, an analysis unit 282, and a determination unit 283.

The memory unit 281 is a component in which regression equations each representing the relationship between a plurality of parameter values and a plurality of candidate compositions of the resin layer 12 are stored in advance. The memory unit 281 is, for example, a memory such as a ROM or a RAM.

In the example of FIG. 2, the memory unit 281 includes a standard regression equation. The standard regression equation represents the relationship between multiple parameter values and a first candidate composition and a second candidate composition at a standard temperature.

The standard regression equation will be described. When the standard regression equation is applied to a plurality of parameter values calculated when the resin layer 12 containing the first candidate composition or the second candidate composition is irradiated with near-infrared rays at a standard temperature, the first similarity and the second similarity are calculated. The first similarity is an index representing whether the resin layer 12 is similar to the first candidate composition. The second similarity is an index representing whether the resin layer 12 is similar to the second candidate composition. When the resin layer 12 contains the first candidate composition, the first similarity is closer to 1 than 0, and the second similarity is closer to 0 than 1. For example, when the resin layer 12 contains the first candidate composition, the first similarity is greater than 0.5, and the second similarity is less than 0.5. In this way, when the resin layer 12 is irradiated with near-infrared rays at a standard temperature, the composition contained in the resin layer 12 can be determined based on whether the similarity is greater than 0.5. The thresholds used in analysis at standard temperatures are also called standard thresholds.

However, when the ambient temperature changes, the target reflection spectrum may change. For example, when the ambient temperature changes, the position of the peak of the target reflection spectrum caused by the first candidate composition or the second candidate composition may change. As a result, the similarity calculated by applying the standard regression equation to the multiple parameter values of the target reflection spectrum also changes. For this reason, when near-infrared rays are irradiated onto the resin layer 12 at a temperature different from the standard temperature, an erroneous determination may be made if the composition of the resin layer 12 is determined using a standard threshold value.

Considering such problems, in this embodiment, it is proposed to use a threshold value corresponding to the ambient temperature when making a judgment based on the similarity. For example, as shown in FIG. 2, it is proposed to store a first threshold value, a second threshold value, a third threshold value, and a fourth threshold value in the memory unit 281. The first threshold value is a threshold value for judging whether the composition of the resin layer 12 is the first candidate composition or the second candidate composition at the first temperature. Similarly, the second threshold value to the fourth threshold value are threshold values for judging whether the composition of the resin layer 12 is the first candidate composition or the second candidate composition at the second temperature to the fourth temperature. In this way, by using different threshold values depending on the ambient temperature, it is possible to make an appropriate judgment regardless of the temperature. The first temperature, the second temperature, the third temperature, and the fourth temperature are, for example, 5° C., 15° C., 25° C., and 35° C.

The analysis unit 282 calculates the similarity by applying a standard regression equation to the multiple parameter values described above. The analysis unit 282 is, for example, a CPU.

The determination unit 283 determines whether the composition of the resin layer 12 is the first candidate composition or the second candidate composition based on a comparison between the similarity calculated by the analysis unit 282 and a threshold value corresponding to the ambient temperature. For example, the determination unit 283 selects a threshold value corresponding to the ambient temperature from among a plurality of threshold values based on information from the temperature monitor 23. For example, when the ambient temperature is the second temperature, the determination unit 283 selects the second threshold value. When the ambient temperature is a value between the second temperature and the third temperature and is closer to the third temperature than the second temperature, the determination unit 283 selects the third threshold value. For example, when the ambient temperature is 22° C., the determination unit 283 selects the third threshold value. Next, the determination unit 283 determines the composition of the resin layer 12 based on whether the similarity is greater than the threshold value. For example, when the second similarity is greater than the selected threshold value, the determination unit 283 determines that the resin layer 12 contains the second candidate composition. The determination unit 283 is, for example, a CPU.

Although not shown, the determination device 28 may have a display unit that displays the determination result. The manner in which the determination result is displayed on the display unit is not particularly limited as long as the determination result can be understood from the display. For example, the similarity calculated in the analysis unit 282 may be displayed. The threshold value used in the determination unit 283 may be displayed. The name of the candidate composition determined as the composition of the resin layer 12 in the determination unit 283 may be displayed.

Next, a specific structure of the inspection system 15 will be described. As shown in FIG. 4, the inspection system 15 may include a case 40 that houses the spectroscopic module 20. The spectroscopic module 20 includes an irradiation device 21 and a detection device 24. The case 40 preferably has a portable form. The weight of the case 40 that houses the irradiation device 21 and the detection device 24 is, for example, 5 kg or less, may be 2 kg or less, and is preferably 1 kg or less. By carrying the case 40, it is possible to inspect objects 10 located in various locations.

The case 40 may include a first surface 41, a second surface 42, and a side surface. The first surface 41 faces the object 10 when the inspection method using the inspection system 15 is performed. The first surface 41 includes a window 48 through which light passes. The light is light emitted from the irradiation device 21 or reflected light received by the detection device 24. The window 48 may include an opening formed in the first surface 41. The second surface 42 faces the first surface 41. The side surface is located between the first surface 41 and the second surface 42. The side surface may include, for example, a first side surface 43, a second side surface 44, a third side surface 45, and a fourth side surface 46. The second surface 42 faces the first surface 41. The second side surface 44 faces the first side surface 43. The first side surface 43 and the second side surface 44 are located between the first surface 41 and the second surface 42.

The case 40 may include, for example, a first outer edge 411, a second outer edge 412, a third outer edge 413, and a fourth outer edge 414. The first outer edge 411, the second outer edge 412, the third outer edge 413, and the fourth outer edge 414 form the outer edge of the portion of the case 40 that faces the object 10.

A part or all of the first outer edge 411, the second outer edge 412, the third outer edge 413, and the fourth outer edge 414 may be the outer edge of the first surface 41. A part or all of the first outer edge 411, the second outer edge 412, the third outer edge 413, and the fourth outer edge 414 may be the outer edge of a side surface such as the first side surface 43, the second side surface 44, the third side surface 45, and the fourth side surface 46.

The dimension K1 of the first outer edge 411 is, for example, 500 mm or less, may be 400 mm or less, or may be 300 mm or less. The dimension K1 of the first outer edge 411 is, for example, 50 mm or more, may be 100 mm or more, or may be 150 mm or more.
The dimension K2 of the third outer edge 413 is, for example, 500 mm or less, may be 400 mm or less, or may be 300 mm or less. The dimension K2 of the third outer edge 413 is, for example, 50 mm or more, may be 100 mm or more, or may be 150 mm or more.
The distance K3 between the first surface 41 and the second surface 42 is, for example, 500 mm or less, may be 400 mm or less, or may be 300 mm or less. The distance K3 is, for example, 50 mm or more, may be 100 mm or more, or may be 150 mm or more.

Figure 5 is a diagram showing an example of how the inspection system 15 is used. In the example shown in Figure 5, the case 40 is held by a person's right hand 61 and left hand 62. The first surface 41 of the case 40 faces the object 10. When performing an inspection method using the inspection system 15, some or all of the first outer edge 411, the second outer edge 412, the third outer edge 413, and the fourth outer edge 414 may be in contact with the object 10.

[Method of manufacturing the inspection system]
Next, a method for manufacturing the above-mentioned inspection system 15 will be described. Specifically, a method for creating the above-mentioned standard regression equation and threshold value will be described.

First, a system including the irradiation device 21, temperature monitor 23, detection device 24, and processing device 27 shown in FIG. 2 is prepared. A first standard sample and a second standard sample are also prepared. The first standard sample includes a first candidate composition. The first candidate composition is, for example, polyethylene (hereinafter also referred to as PE). The second standard sample includes a second candidate composition. The second candidate composition is, for example, ethylene-vinyl acetate copolymer resin (hereinafter also referred to as EVA).

Next, a standard sample irradiation step is performed in which the first standard sample and the second standard sample are irradiated with light L1 using the irradiation device 21 at multiple temperatures including the standard temperature. For example, near-infrared light is irradiated to each of the first standard sample and the second standard sample at each of the standard temperature, the first temperature, the second temperature, the third temperature, and the fourth temperature. Next, a standard sample detection step is performed in which standard reflection spectra from the first standard sample and the second standard sample are obtained using the detection device 24 and the generation device 25 at multiple temperatures including the standard temperature. The standard reflection spectrum from the first standard sample is also referred to as the first standard reflection spectrum. The standard reflection spectrum from the second standard sample is also referred to as the second standard reflection spectrum. Figure 6 is a diagram showing an example of the second standard reflection spectrum. In the example shown in Figure 6, the standard reflection spectrum is an absorbance spectrum.

Then, using the processing device 27, a standard sample processing step is performed at multiple temperatures including the standard temperature to perform a predetermined process on the first standard reflectance spectrum and the second standard reflectance spectrum. This causes the parameter value to be calculated at each wavelength point. As described above, the process includes, for example, averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof. The process may be performed on the values of the reflected light intensity at multiple wavelength points of the first standard reflectance spectrum and the second standard reflectance spectrum. Alternatively, the process may be performed on a spectrum containing information equivalent to that of the first standard reflectance spectrum and the second standard reflectance spectrum, such as an absorbance spectrum.

The parameter value at each wavelength point calculated from the first standard reflectance spectrum is also referred to as the first standard parameter value. The parameter value at each wavelength point calculated from the second standard reflectance spectrum is also referred to as the second standard parameter value. The first standard parameter value and the second standard parameter value are each calculated for each temperature.

Then, the processing device 27 is used to perform a regression equation creation process. In the regression equation creation process, the multiple first standard parameter values and multiple second standard parameter values obtained at the standard temperature are analyzed using a multivariate analysis method. The multivariate analysis is performed, for example, by recording a program for multivariate analysis in the memory unit 281 and executing this program in the analysis unit 282. In this way, a standard regression equation is created. The standard regression equation represents the relationship that holds at the standard temperature between the parameter values at multiple wavelength points and the first and second candidate compositions.

Figure 7 is a diagram showing the results of analyzing a plurality of first standard parameter values and a plurality of second standard parameter values using a multivariate analysis method. In the example shown in Figure 7, PLS discriminant analysis (PLS-DA) was adopted as the multivariate analysis. As the software for performing PLS discriminant analysis (PLS-DA), the multivariate analysis software The Unscrambler X made by Kamo Software was used. The "Y predicted value" shown on the vertical axis of Figure 7 corresponds to the similarity to the candidate composition Y in a situation where two candidate compositions X and Y exist. When the candidate composition X corresponds to the first candidate composition and the candidate composition Y corresponds to the second candidate composition, the Y predicted value corresponds to the second similarity. The "Y reference value" shown on the horizontal axis of Figure 7 indicates whether the composition of the standard sample is the candidate composition Y. In the first standard sample (= candidate composition X), the Y reference value = 0. In the second standard sample (= candidate composition Y), the Y reference value = 1. In FIG. 7, the multiple points surrounded by a dotted line labeled M1 represent the second similarity calculated from the test results of the multiple first standard samples at the standard temperature. In FIG. 7, the multiple points surrounded by a dotted line labeled M2 represent the second similarity calculated from the test results of the multiple second standard samples at the standard temperature. A standard regression equation is created so that the predicted Y values of the multiple points located in the range of M1 approach 0 and the predicted Y values of the multiple points located in the range of M2 approach 1. The created standard regression equation is recorded in the memory unit 281.

In the regression equation creation process, multivariate analysis other than PLS discriminant analysis (PLS-DA) may be used. For example, multivariate analysis methods such as cluster analysis, support vector machine, KNN (K-Nearest Neighbor), and SIMCA (Soft Independent Modeling of Class Analogy) may be used.

Next, a threshold creation process is performed. In the threshold creation process, a standard regression equation is applied to a plurality of first standard parameter values and a plurality of second standard parameter values at temperatures other than the standard temperature. As a result, the first similarity and second similarity of the first standard sample and the first similarity and second similarity of the second standard sample are calculated at each temperature other than the standard temperature. Figures 8 to 11 are diagrams showing the results of applying the standard regression equation to the first standard parameter values and the second standard parameter values at the first to fourth temperatures. As can be seen from Figures 8 to 11, the average values of the first similarity and second similarity of the first standard sample and the average values of the first similarity and second similarity of the second standard sample vary depending on the temperature.

In the threshold creation process, a threshold for distinguishing between the first candidate composition and the second candidate composition is calculated for each temperature based on the first similarity and second similarity of the first standard sample and the first similarity and second similarity of the second standard sample at each temperature. FIG. 12 is a diagram showing the calculation results of the thresholds at the first to fourth temperatures. FIG. 12 is obtained by projecting the calculation results of the similarity at each temperature, as shown in FIGS. 8 to 11, onto the vertical axis "second similarity". In FIG. 12, the symbols TH1 to TH4 represent the first to fourth thresholds at the first to fourth temperatures, respectively. The created first to fourth thresholds are recorded in the memory unit 281.

The specific method of calculating the thresholds at the first to fourth temperatures is arbitrary. For example, at each temperature, the average or median of the second similarity for multiple first standard samples and the average or median of the second similarity for multiple second standard samples are calculated. The average of the two average values or the average of the two medians can be used as the threshold for the second similarity.

[Testing method]
Next, a method for inspecting the object 10 using the inspection system 15 will be described. First, the inspection system 15 is prepared, which includes the irradiation device 21, the temperature monitor 23, the detection device 24, the processing device 27, and the determination device 28 shown in Fig. 2. Also, the object 10 to be used as a sample is prepared. In this embodiment, the sample includes either the first candidate composition or the second candidate composition.

The inspection method includes a temperature monitoring step of monitoring the ambient temperature, and an inspection step of inspecting the object 10. The inspection method may also include a calibration step of calibrating the inspection system 15 using a calibration sample 51.

[Inspection process]
The inspection process will be described. The inspection process includes a first irradiation process, a first detection process, a first generation process, and a first processing process. In the first irradiation process, the irradiation device 21 irradiates the object 10 with light L1. In the first detection process, the detection device 24 detects the intensity of the light L2 reflected by the object 10. In the first generation process, the first generation unit 251 of the generation device 25 generates a sample reflection spectrum. The sample reflection spectrum includes information on the intensity of the light L2 at a plurality of wavelength points. The information may be the intensity of the light L2 itself. The information may be information calculated by performing some processing on the intensity of the light L2, such as absorbance.

Then, the first processing unit 271 of the processing device 27 performs a first processing step to process the sample reflectance spectrum. This causes parameter values to be calculated at each wavelength point. The processing performed on the sample reflectance spectrum is the same as the processing performed on the standard reflectance spectrum.

Then, the analysis unit 282 of the determination device 28 performs an analysis step in which a standard regression equation is applied to multiple parameter values of the sample reflection spectrum. This allows the similarity to be calculated. For example, a second similarity for the second candidate composition is calculated.

Then, the judgment unit 283 of the judgment device 28 performs the judgment process. The judgment unit 283 first selects a threshold value corresponding to the ambient temperature from among a plurality of threshold values. For example, if the second temperature among the first to fourth temperatures is closest to the ambient temperature, the judgment unit 283 selects the second threshold value. The judgment unit 283 then judges whether the composition of the resin layer 12 is the first candidate composition or the second candidate composition based on a comparison between the similarity calculated by the analysis unit 282 and the second threshold value. For example, if the second similarity calculated by the analysis unit 282 is greater than the second threshold value, the judgment unit 283 judges that the composition of the resin layer 12 is the second candidate composition.

Figure 13 shows the results of determining the compositions of multiple samples at multiple temperatures. In Figure 13, a circular marker indicates the second similarity of a sample containing a first candidate composition, and a triangular marker indicates the second similarity of a sample containing a second candidate composition. When a marker is located below the thresholds TH1 to TH4, it means that the composition of the sample has been determined to be the first candidate composition. When a marker is located above the thresholds TH1 to TH4, it means that the composition of the sample has been determined to be the second candidate composition.

[Calibration method]
Next, the calibration process will be described. The calibration process includes a second irradiation process, a second detection process, and a second generation process. In the second irradiation process, the irradiation device 21 irradiates the calibration sample 51 with light L1. In the second detection process, the detection device 24 detects the intensity of the light L2 reflected by the calibration sample 51. In the second generation process, the second generation unit 252 of the generation device 25 generates a calibration spectrum. The calibration spectrum includes information on the intensity of the light L2 at a plurality of wavelength points. The information may be the intensity of the light L2 itself. FIG. 14 is a diagram showing an example of a calibration spectrum.

As shown in FIG. 15, the calibration process may be performed with the calibration sample 51 in contact with the first surface 41 so as to cover the window 48. In this case, the calibration sample 51 may be fixed to the case 40. For example, the calibration sample 51 may be pressed against the window 48 by a rubber band 52 wrapped around the case 40. Although not shown, the calibration sample 51 may be separated from the first surface 41. Note that, when the calibration sample 51 is separated from the first surface 41, it is preferable to make the surroundings a dark room. This can suppress the inspection system 15 from detecting external light. In addition, in the process of inspecting the object 10, it is preferable that the distance between the object 10 and the first surface 41 is the same as the distance between the standard sample 51 and the first surface 41.

The timing of performing the calibration process is arbitrary. For example, the calibration process may be performed before starting the inspection process. Alternatively, the calibration process may be performed between multiple inspection processes.

[Effects of this embodiment]
According to this embodiment, the accuracy of the composition determination can be improved by using an appropriate threshold value depending on the ambient temperature.

The inspection system 15 and inspection method according to the present embodiment may be used in combination with a conventional inspection system and inspection method that determines the composition based on the peak wavelength of the spectrum. For example, consider the case of determining whether the object 10 is a first candidate composition, a second candidate composition, or a third candidate composition. Assume that the peak wavelength of the third candidate composition is significantly different from the peak wavelength of the first candidate composition and the peak wavelength of the second candidate composition. In this case, it can be determined whether the object 10 is a group including the first candidate composition and the second candidate composition, or a third candidate composition, based on the position of the peak wavelength of the reflection spectrum from the object. If the object 10 is determined to be a group including the first candidate composition and the second candidate composition, then the inspection system 15 and inspection method according to the present embodiment are used to determine whether the object 10 is a first candidate composition or a second candidate composition.

(Modification of the first embodiment)
By using the above-described inspection system 15 and inspection method, it is also possible to determine the composition of the object 10 in more detail. For example, the inspection system 15 and inspection method can inspect the object 10 determined as the second candidate composition in more detail. For example, the inspection system 15 and inspection method can determine whether the object 10 is the 21st candidate composition or the 22nd candidate composition. The 21st candidate composition and the 22nd candidate composition are compositions that are subordinate to the second candidate composition or similar to the second candidate composition.
The combination of the first candidate composition, the second candidate composition, the twenty-first candidate composition, and the twenty-second candidate composition is arbitrary. For example, the first candidate composition may be polyvinyl chloride (hereinafter also referred to as PVC), the twenty-first candidate composition may be PE, and the twenty-second candidate composition may be EVA. In this case, the second candidate composition is a group including PE and EVA. In the inspection method, first, it is determined whether the object 10 is PVC or a group including PE and EVA. If it is determined that the object 10 is a group including PE and EVA, it is then determined whether the object 10 is PE or EVA.

In this modified example, the memory unit 281 stores a sub-standard regression equation that represents the relationship between the parameter value at standard temperature and the 21st and 22nd candidate compositions. If the composition of the resin layer 12 is determined to be the second candidate composition, the analysis unit 282 applies the sub-standard regression equation to the multiple parameter values of the sample. This calculates the sub-similarity. The determination unit 283 determines whether the composition of the resin layer is the 21st or 22nd candidate composition based on a comparison between the sub-similarity and a sub-threshold value.

As in the above embodiment, the sub-thresholds are stored in the storage unit 281 for each temperature. For example, the storage unit 281 stores a first sub-threshold, a second sub-threshold, a third sub-threshold, and a fourth sub-threshold. The first sub-threshold is a threshold for determining whether the composition of the resin layer 12 is the 21st candidate composition or the 22nd candidate composition at the first temperature. Similarly, the second to fourth sub-thresholds are thresholds for determining whether the composition of the resin layer 12 is the 21st candidate composition or the 22nd candidate composition at the second to fourth temperatures.

The method for creating the sub-standard regression equation and the first to fourth sub-thresholds is similar to the method for creating the standard regression equation and the first to fourth thresholds, so a detailed explanation is omitted.

The embodiment described above can be modified in various ways. Other embodiments will be described below with reference to the drawings as necessary. In the following description and the drawings used in the following description, the same reference numerals are used for parts that can be configured similarly to the embodiment described above as for the corresponding parts in the embodiment described above. Duplicate descriptions will be omitted. Furthermore, if it is clear that the effects obtained in the embodiment described above can also be obtained in other embodiments, the description may be omitted.

Second Embodiment
In the first embodiment, an example in which different threshold values are used depending on the ambient temperature is described. In the second embodiment, an example in which different regression equations are used depending on the ambient temperature is described.

Figure 16 is a block diagram showing an inspection system 15 for inspecting an object 10. A memory unit 281 stores a first regression equation, a second regression equation, a third regression equation, a fourth regression equation, and a standard threshold value. The first regression equation represents the relationship at the first temperature between the parameter value at each wavelength point and the first and second candidate compositions. Similarly, the second to fourth regression equations represent the relationship at the second to fourth temperatures between the parameter value at each wavelength point and the first and second candidate compositions. The standard threshold value is a threshold value for determining whether the composition of the resin layer 12 is the first or second candidate composition at multiple temperatures. For example, the standard threshold value can be used at any of the first to fourth temperatures.

[Method of manufacturing the inspection system]
A method for manufacturing the inspection system 15 will be described. First, similarly to the first embodiment, a standard sample irradiation step, a standard sample detection step, and a standard sample processing step are performed. As a result, a first standard reflectance spectrum and a second standard reflectance spectrum are calculated for each temperature.

Then, the processing device 27 is used to perform a regression equation creation process. In the regression equation creation process, the multiple first standard parameter values and multiple second standard parameter values are analyzed for each temperature using a multivariate analysis method. For example, the multiple first standard parameter values and multiple second standard parameter values calculated at the first temperature are analyzed by the multivariate analysis method. This creates the first regression equation. Similarly, the multiple first standard parameter values and multiple second standard parameter values calculated at the second to fourth temperatures are each analyzed by the multivariate analysis method. This creates the second to fourth regression equations. Each of the first to fourth regression equations is configured to distinguish the first standard sample from the second standard sample with the standard threshold as the boundary. The standard threshold is, for example, 0.5. The created first to fourth regression equations are recorded in the memory unit 281.

[Testing method]
A method for inspecting the object 10 using the inspection system 15 will be described. First, the first irradiation step, the first detection step, the first generation step, and the first processing step are performed in the same manner as in the first embodiment. As a result, parameter values of the sample at each wavelength point are calculated.

Then, the analysis unit 282 of the determination device 28 performs the analysis step. The analysis unit 282 first selects a regression equation corresponding to the ambient temperature from among the multiple regression equations. For example, if the second temperature among the first to fourth temperatures is closest to the ambient temperature, the analysis unit 282 selects the second regression equation. The analysis unit 282 then applies the second regression equation to the multiple parameter values of the sample. This calculates the similarity of the sample. For example, the second similarity to the second candidate composition is calculated.

Then, the judgment unit 283 of the judgment device 28 performs a judgment process. The judgment unit 283 judges whether the composition of the resin layer 12 is the first candidate composition or the second candidate composition based on a comparison between the similarity calculated by the analysis unit 282 and the standard threshold value. For example, if the second similarity calculated by the analysis unit 282 is greater than the standard threshold value, the judgment unit 283 judges that the composition of the resin layer 12 is the second candidate composition.

[Effects of this embodiment]
According to this embodiment, by using an appropriate regression equation depending on the ambient temperature, it is possible to improve the accuracy of the composition determination.

(Modification of the second embodiment)
By using the above-described inspection system 15 and inspection method, it is also possible to determine the composition of the object 10 in more detail. For example, the inspection system 15 and inspection method can inspect the object 10 determined as the second candidate composition in more detail. For example, the inspection system 15 and inspection method can determine whether the object 10 is the 21st candidate composition or the 22nd candidate composition. The 21st candidate composition and the 22nd candidate composition are compositions that are subordinate to the second candidate composition or similar to the second candidate composition.

In this modified example, the memory unit 281 stores a sub-regression equation that represents the relationship between the parameter value and the 21st and 22nd candidate compositions. If the composition of the resin layer 12 is determined to be the second candidate composition, the analysis unit 282 applies the sub-regression equation to the multiple parameter values of the sample. This calculates a sub-similarity. The determination unit 283 determines whether the composition of the resin layer is the 21st or 22nd candidate composition based on a comparison between the sub-similarity and the sub-standard threshold.

As in the above-described embodiment, the sub-regression equations are stored in the storage unit 281 for each temperature. For example, the storage unit 281 stores a first sub-regression equation, a second sub-regression equation, a third sub-regression equation, and a fourth sub-regression equation. The first sub-regression equation represents the relationship between the parameter value at each wavelength point and the 21st and 22nd candidate compositions at the first temperature. Similarly, the second to fourth sub-regression equations represent the relationship between the parameter value at each wavelength point and the 21st and 22nd candidate compositions at the second to fourth temperatures. The sub-standard threshold is a threshold for determining whether the composition of the resin layer 12 is the 21st or 22nd candidate composition at multiple temperatures. For example, the sub-standard threshold can be used at any of the first to fourth temperatures. The sub-standard threshold is, for example, 0.5.

The method for creating the first to fourth sub-regression equations is similar to the method for creating the first to fourth regression equations, so a detailed explanation will be omitted.

Third Embodiment
In the first and second embodiments, examples have been described in which different thresholds or regression equations are used depending on the ambient temperature. In the third embodiment, an example will be described in which one general-purpose regression equation and one general-purpose threshold are used.

Figure 17 is a block diagram showing an inspection system 15 for inspecting an object 10. A memory unit 281 stores a general regression equation and a general threshold. The general regression equation represents a relationship between a parameter value at each wavelength point and a first candidate composition and a second candidate composition that is valid at any of a plurality of temperatures. For example, the general regression equation can be used at any of the first to fourth temperatures. The general threshold is used in combination with the general regression equation. The general threshold is a threshold for determining whether the composition of the resin layer 12 is the first candidate composition or the second candidate composition at multiple temperatures. For example, the general threshold can be used at any of the first to fourth temperatures. The general threshold is, for example, 0.5.

[Method of manufacturing the inspection system]
A method for manufacturing the inspection system 15 will be described. First, similarly to the first embodiment, a standard sample irradiation step, a standard sample detection step, and a standard sample processing step are performed. As a result, a first standard reflection spectrum and a second standard reflection spectrum are calculated for each temperature. For example, the first standard reflection spectrum and the second standard reflection spectrum at the first temperature, the second temperature, the third temperature, and the fourth temperature are calculated, respectively.

Next, a regression equation creation process is performed using the processing device 27. In the regression equation creation process, the first standard parameter value and the second standard parameter value at each wavelength point calculated at multiple temperatures are analyzed together using a multivariate analysis method. FIG. 18 is a diagram showing the calculation result of the general regression equation. In FIG. 18, a plurality of points surrounded by a dotted line with a symbol M3 represent the second similarity calculated from the test results at multiple temperatures of multiple first standard samples. Since M3 includes the test results at multiple temperatures, the range of M3 in the vertical axis direction is wider than the range of M1 in FIG. 7. In FIG. 18, a plurality of points surrounded by a dotted line with a symbol M4 represent the second similarity calculated from the test results at multiple temperatures of multiple second standard samples. Since M4 includes the test results at multiple temperatures, the range of M4 in the vertical axis direction is wider than the range of M2 in FIG. 7. A generalized regression equation is created so that the predicted Y values of multiple points located in the range of M3 approach 0 and the predicted Y values of multiple points located in the range of M4 approach 1. The generalized threshold is set so that the first standard sample and the second standard sample can be distinguished. The generalized threshold is, for example, 0.5. The created generalized regression equation and generalized threshold are recorded in the memory unit 281.

[Testing method]
A method for inspecting the object 10 using the inspection system 15 will be described. First, the first irradiation step, the first detection step, the first generation step, and the first processing step are performed in the same manner as in the first embodiment. As a result, parameter values of the sample at each wavelength point are calculated.

Then, the analysis unit 282 of the determination device 28 performs an analysis process. The analysis unit 282 applies a general regression equation to multiple parameter values of the sample. This calculates the similarity of the sample. For example, a second similarity to a second candidate composition is calculated.

Then, the judgment unit 283 of the judgment device 28 performs a judgment process. The judgment unit 283 judges whether the composition of the resin layer 12 is the first candidate composition or the second candidate composition based on a comparison between the similarity calculated by the analysis unit 282 and the general-purpose threshold value. For example, if the second similarity calculated by the analysis unit 282 is greater than the general-purpose threshold value, the judgment unit 283 judges that the composition of the resin layer 12 is the second candidate composition.

Figure 19 shows the results of determining the compositions of multiple samples at multiple temperatures. In Figure 19, a circular marker indicates the second similarity of a sample containing a first candidate composition, and a triangular marker indicates the second similarity of a sample containing a second candidate composition. When a marker is located below the generic threshold TH10, it means that the composition of the sample has been determined to be the first candidate composition. When a marker is located above the generic threshold TH10, it means that the composition of the sample has been determined to be the second candidate composition.

[Effects of this embodiment]
According to this embodiment, the accuracy of composition determination can be improved by using a general-purpose regression equation and a general-purpose threshold value that reflect the inspection results of the standard sample at a plurality of temperatures.

(Modification of the third embodiment)
By using the above-described inspection system 15 and inspection method, it is also possible to determine the composition of the object 10 in more detail. For example, the inspection system 15 and inspection method can inspect the object 10 determined as the second candidate composition in more detail. For example, the inspection system 15 and inspection method can determine whether the object 10 is the 21st candidate composition or the 22nd candidate composition. The 21st candidate composition and the 22nd candidate composition are compositions that are subordinate to the second candidate composition or similar to the second candidate composition.

In this modified example, the memory unit 281 stores a sub-general regression equation that represents the relationship between the parameter value and the 21st and 22nd candidate compositions. If the composition of the resin layer 12 is determined to be the second candidate composition, the analysis unit 282 applies the sub-general regression equation to the multiple parameter values of the sample. This calculates a sub-similarity. The determination unit 283 determines whether the composition of the resin layer is the 21st or 22nd candidate composition based on a comparison between the sub-similarity and a sub-general threshold.

As in the above embodiment, the sub-general regression equation represents a relationship between the parameter values at each wavelength point and the 21st and 22nd candidate compositions that is valid at any of a plurality of temperatures. The sub-general threshold is used in combination with the sub-general regression equation. The sub-general threshold is a threshold for determining whether the composition of the resin layer 12 is the 21st or 22nd candidate composition at a plurality of temperatures. For example, the sub-general threshold can be used at any of the first to fourth temperatures. The sub-general threshold is, for example, 0.5.

The method for creating a sub-general regression equation is the same as the method for creating a general regression equation, so a detailed explanation will be omitted.

(Fourth embodiment)
In the fourth embodiment, an example of eliminating a sample reflection spectrum containing a lot of noise will be described. The configuration of the inspection system 15 of the fourth embodiment is the same as the configuration of the inspection system 15 of the first embodiment shown in FIG.

[Testing method]
A method for inspecting the object 10 using the inspection system 15 will be described. Fig. 20 is a flow chart showing an example of the inspection process. First, a first irradiation process, a first detection process, and a first generation process are performed in the same manner as in the first embodiment. As a result, a reflection spectrum of the sample is generated, as shown by reference symbol S11 in Fig. 20.

The sample reflection spectrum includes information about the intensity of light L2 at multiple wavelength points. The information may be the intensity of light L2 itself. The information may be calculated by performing some processing on the intensity of light L2, such as absorbance. Figure 21 shows an example of a sample reflection spectrum S1 that includes absorbance at multiple wavelength points.

Then, the first processing unit 271 of the processing device 27 performs a first processing step to process the sample reflectance spectrum. The first processing step includes a local averaging step, a first noise evaluation step, and a first recording step.

In the local averaging process, the target reflectance spectrum is locally averaged, as shown by reference symbol S12 in FIG. 20. For example, a process of calculating a local average value at each wavelength point of the sample reflectance spectrum is performed. The local average value is calculated by averaging the value of the sample reflectance spectrum at the target wavelength point and the values of the sample reflectance spectrum at multiple wavelength points located in the vicinity of the target wavelength point.

An example will be described in which the value of the sample reflectance spectrum is absorbance. For example, the local average value at a wavelength point of 1200 nm is calculated as follows:
First, the absorbance Abs (1200) at 1200 nm is obtained. Then, the values of the sample reflection spectrum at a plurality of wavelength points located in the vicinity of 1200 nm are obtained. When the resolution of the detection device 24 is 3 nm, for example, the absorbance Abs (1188) at 1188 nm, the absorbance Abs (1191) at 1191 nm, the absorbance Abs (1194) at 1194 nm, the absorbance Abs (1197) at 1197 nm, the absorbance Abs (1203) at 1203 nm, the absorbance Abs (1206) at 1206 nm, the absorbance Abs (1209) at 1209 nm, and the absorbance Abs (1212) at 1212 nm are obtained. Then, the average value of the absorbance at these nine wavelength points is calculated. This average value is the local average value at the wavelength point of 1200 nm.

The local average value is calculated at each wavelength point. This allows the generation of a locally averaged spectrum S2, as shown in FIG. 22. In FIG. 22, the sample reflectance spectrum S1 is represented by a dashed line, and the averaged spectrum S2 is represented by a solid line.

The number of wavelength points when calculating the local average value is not limited to nine as described above. The number of wavelength points may be, for example, five or more, and may be seven or more. The number of wavelength points may be, for example, 15 or less, and may be 13 or less.

In the first noise evaluation process, as shown by symbol S13 in FIG. 20, the difference between the sample reflection spectrum S1 and the locally averaged spectrum S2 is calculated. For example, at each wavelength point, the locally averaged value of the averaged spectrum S2 is subtracted from the absorbance of the sample reflection spectrum S1. The spectrum generated in this way is also called the difference spectrum. FIG. 23 is a diagram showing an example of the difference spectrum S3. In FIG. 23, the sample reflection spectrum S1 is represented by a dashed line, the averaged spectrum S2 is represented by a dashed line, and the difference spectrum S3 is represented by a solid line.

Next, in the first noise evaluation process, the standard deviation of the difference between the sample reflection spectrum S1 and the locally averaged spectrum S2 is calculated. The standard deviation represents the degree of variation in the difference between the sample reflection spectrum S1 and the locally averaged spectrum S2. If the target reflection spectrum S1 does not contain noise, the difference between the sample reflection spectrum S1 and the locally averaged spectrum S2 is expected to be almost zero. Therefore, a small standard deviation means that the sample reflection spectrum S1 contains little noise. A large standard deviation means that the sample reflection spectrum S1 contains a lot of noise.

Note that, regardless of noise, the difference between the target reflection spectrum S1 and the locally averaged spectrum S2 is large at the position of the peak of the sample reflection spectrum S1. In order to properly evaluate the noise contained in the sample reflection spectrum S1, it is preferable to calculate the standard deviation of the differences at a position away from the peak of the sample reflection spectrum S1. For example, as shown in FIG. 23, the standard deviation of the differences may be calculated in a stable region R1 that is away from the peak. Specifically, first, difference values at multiple wavelength points contained in the stable region R1 are obtained. Next, the standard deviation of the multiple difference values is calculated.

The stable region R1 may be determined in advance based on information about the material expected to be contained in the object 10. For example, if the object 10 is expected to contain polyethylene, the object reflectance spectrum S1 includes a peak that appears at approximately 1200 nm. In this case, the wavelength range from 970 nm to 1100 nm may be determined in advance as the stable region R1.

Next, in the first noise evaluation process, a first noise index is calculated based on the standard deviation. The first noise index is an index of the degree of noise contained in the sample reflection spectrum S1. The first noise index is, for example, the reciprocal of the standard deviation. A large first noise index means that the sample reflection spectrum S1 contains less noise. A small first noise index means that the sample reflection spectrum S1 contains more noise. The calculated first noise index may be displayed on a display of a computer constituting the processing device 27.

Next, in the first noise evaluation step, the first noise index is compared with the first noise threshold TH21, as shown by reference symbol S14 in FIG. 20. If the first noise index is equal to or greater than the first noise threshold TH21, the first processing unit 271 performs the first recording step. In the first recording step, a sample reflection spectrum S1 having a first noise index equal to or greater than the first noise threshold TH21 is recorded, as shown by reference symbol S15 in FIG. 20. The recorded sample reflection spectrum S1 may be used in the determination device 28. The comparison result between the first noise index and the first noise threshold TH21 may be displayed on the display of the computer constituting the processing device 27.

If the first noise index is less than the first noise threshold TH21, steps S11, S12, S13, and S14 may be performed again as shown in FIG. 20. That is, the first irradiation step, the first detection step, the first generation step, and the first processing step may be performed again on the object 10. When the case 40 is held by hand as shown in FIG. 5, it is considered that noise may occur in the sample reflection spectrum S1 due to hand movement during the inspection step. According to the inspection step shown in FIG. 20, it is possible to eliminate the sample reflection spectrum S1 that contains a lot of noise caused by hand movement, etc. In addition, the inspection step can be repeatedly performed until a sample reflection spectrum S1 having a first noise index equal to or greater than the first noise threshold TH21 is obtained. Therefore, an appropriate sample reflection spectrum S1 can be recorded.

Then, the first processing unit 271 performs a predetermined process on the recorded sample reflection spectrum S1. This causes the parameter values to be calculated at each wavelength point. The process is the same as in the first embodiment described above.

Next, as in the first embodiment described above, the analysis unit 282 performs the analysis process, and the determination unit 283 performs the determination process. This allows the composition of the resin layer 12 to be determined.

[Effects of this embodiment]
According to this embodiment, the noise contained in the sample reflection spectrum can be evaluated. Therefore, it is possible to eliminate a sample reflection spectrum that contains a large amount of noise caused by hand movement or the like. In addition, the inspection process can be repeated until a sample reflection spectrum having a first noise index equal to or greater than the first noise threshold value TH21 is obtained. Therefore, an appropriate sample reflection spectrum can be recorded. Therefore, the composition of the material constituting the resin layer 12 can be determined more appropriately.

(First Modification of the Fourth Embodiment)
The inspection process in the above-mentioned fourth embodiment may be applied to the standard reflection spectrum. That is, the noise contained in the standard reflection spectrum may be evaluated by the above-mentioned inspection process. Furthermore, a standard reflection spectrum containing a large amount of noise may be eliminated. Furthermore, the standard sample irradiation process and the standard sample detection process may be repeatedly performed until a standard reflection spectrum having a first noise index equal to or greater than the first noise threshold value TH21 is obtained. This allows an appropriate standard reflection spectrum to be recorded. Therefore, the standard regression equation and the first to fourth threshold values can be created more appropriately.

(Second Modification of the Fourth Embodiment)
The inspection process in the above-mentioned fourth embodiment may be combined with the second embodiment. That is, the noise contained in the sample reflection spectrum or the standard reflection spectrum in the second embodiment may be evaluated by the above-mentioned inspection process. This makes it possible to record an appropriate sample reflection spectrum or standard reflection spectrum. This makes it possible to more appropriately create, for example, the first regression equation to the fourth regression equation.

(Third Modification of the Fourth Embodiment)
The inspection process in the above-mentioned fourth embodiment may be combined with the third embodiment. That is, the noise contained in the sample reflectance spectrum or the standard reflectance spectrum in the third embodiment may be evaluated by the above-mentioned inspection process. This allows an appropriate sample reflectance spectrum or standard reflectance spectrum to be recorded. This allows, for example, a general-purpose regression equation to be created more appropriately.

(Example of the fourth embodiment)
(Example A1)
The object spectrum of the first object 10 was measured using the above-described inspection system 15. The first object 10 has a resin layer 12 containing polyethylene. The first object 10 is attached to a wall. The measurement was performed with the inspection system 15 in contact with the object 10 while being held by a human hand. The measurement result of the object spectrum is shown in FIG. 24. The conditions of the detection process are as follows:
・Wavelength range: 900nm to 1700nm
・Resolution: 3.5nm
Exposure time: 1.27 ms
Number of integrations: 19 The number of integrations is the number of times each wavelength point is scanned in sequence from the lower limit to the upper limit of the wavelength range. When the number of integrations is 19, the detector 24 has 19 detection values at each wavelength point. The detector 24 outputs the average value of the 19 detection values as the light intensity at each wavelength point.

Next, a locally averaged spectrum was generated by locally averaging the target spectrum. The locally averaged spectrum is shown in Figure 24. Next, a difference spectrum was generated based on the difference between the target spectrum and the locally averaged spectrum. The difference spectrum is shown in Figure 24. In Figure 24, as in Figure 23, the target spectrum is represented by a dashed dotted line, the averaged spectrum is represented by a solid line, and the difference spectrum is represented by a solid line.

Next, the standard deviation was calculated based on the value of the difference spectrum in the wavelength range of 970 nm to 1100 nm. Next, the first noise index was calculated based on the standard deviation. The first noise index was 4905. The first noise threshold TH21 was 2500. Therefore, it can be said that the noise contained in the target spectrum shown in FIG. 24 is small.

(Examples A2 to A6)
Using the above-described inspection system 15, object spectra of the second object 10 to the sixth object 10, which are different from the first object 10, were measured. The second object 10 to the sixth object 10 have a resin layer 12 containing polyethylene, similar to the case of Example A1. The measurement conditions are the same as those of Example A1. The measurement results of the object spectra of the second object 10 to the sixth object 10 are shown in Figures 25 to 29, respectively.

Next, a locally averaged spectrum was generated by locally averaging the target spectrum. The locally averaged spectra of the second to sixth objects 10 are shown in Figures 25 to 29, respectively. Next, a difference spectrum was generated based on the difference between the target spectrum and the locally averaged spectrum. The difference spectra of the second to sixth objects 10 are shown in Figures 25 to 29, respectively.

Next, the standard deviation was calculated based on the difference spectrum values in the wavelength range of 970 nm to 1100 nm. Next, the first noise index was calculated based on the standard deviation. The first noise indexes for the second object 10 to the sixth object 10 were 11188, 7405, 2375, 1641, and 1136, respectively.

(Example A7)
The above-described inspection system 15 was used to determine the composition of a first plurality of samples and a second plurality of samples, the first samples including a first candidate composition, and the second samples including a second candidate composition.

The target spectra of the multiple first samples and the multiple second samples were each measured at standard temperature. The first processing unit 271 performed a predetermined process on the target spectra. As a result, parameter values were calculated at each wavelength point. Next, the analysis unit 282 applied a standard regression equation to the parameter values. As a result, the second similarities of the multiple first samples and the multiple second samples were each calculated, as shown in FIG. 30.

As shown by the dotted line marked with the symbol M5 in FIG. 30, the second similarity between the two first samples was greater than the standard threshold value TH0. In other words, the two first samples were erroneously determined to be the second candidate composition.

The first noise indexes of the two incorrectly determined first samples were evaluated. The first noise indexes of the two first samples were 2400 and 1700, respectively.

Fifth embodiment
If the calibration spectrum contains noise, the effect of the noise will also appear in the target spectrum. Therefore, it is preferable to reduce the noise in the calibration spectrum as much as possible. In this embodiment, a method for reducing the noise in the calibration spectrum will be described.

Figure 31 is a block diagram showing an inspection system 15 for inspecting an object 10. As shown in Figure 31, the processing device 27 may include a second processing unit 272. The configuration of the inspection system 15 in Figure 31 is the same as the configuration of the inspection system 15 according to the first embodiment in Figure 2, except that the processing device 27 includes the second processing unit 272.

The second processing unit 272 performs some processing on the calibration spectrum. For example, the second processing unit 272 performs second differentiation to generate a second derivative spectrum by second-order differentiation of the calibration spectrum. The second processing unit 272 also performs a second noise evaluation to calculate a second noise index based on the second derivative spectrum. The second processing unit 272 may perform a second recording to record the calibration spectrum when the second noise index is equal to or greater than a second threshold value.

The calibration process will now be described. The calibration process includes a second irradiation process, a second detection process, a second generation process, and a second processing process. The second processing process processes the calibration spectrum generated in the second generation process.

Figure 32 is a flow chart showing an example of a calibration process. In the second irradiation process, the irradiation device 21 irradiates the calibration sample 51 with light L1. In the second detection process, the detection device 24 detects the intensity of the light L2 reflected by the calibration sample 51. In the second generation process, the second generation unit 252 of the generation device 25 generates a calibration spectrum. The calibration spectrum includes information about the intensity of the light L2 at multiple wavelength points. In this way, the calibration spectrum can be measured, as shown by the symbol S21 in Figure 32.

Then, the second processing unit 272 of the processing device 27 performs a second processing step to process the calibration spectrum. The second processing step includes a second differentiation step, a second noise evaluation step, and a second recording step.

In the second-order differentiation step, the calibration spectrum is second-order differentiated as shown by reference symbol S22 in FIG. 32. This generates a second-order derivative spectrum. A small value of the second-order derivative spectrum means that the calibration spectrum contains less noise.

In the second noise evaluation step, a second noise index is calculated based on the second derivative spectrum. For example, first, as shown by reference symbol S23 in FIG. 32, the standard deviation of the values of the second derivative spectrum at multiple wavelength points is calculated. The standard deviation may be calculated based on the entire wavelength range of the second derivative spectrum, or may be calculated based on a portion of the wavelength range of the second derivative spectrum. For example, as shown in FIG. 14, if it is expected that the value of the calibration spectrum will decrease in the wavelength range of 1600 nm or more, the standard deviation may be calculated without considering the value of the second derivative spectrum in the wavelength range of 1600 nm or more.

Next, in the second noise evaluation process, a second noise index is calculated based on the standard deviation. The second noise index is an index of the degree of noise contained in the calibration spectrum. The second noise index is, for example, the reciprocal of the standard deviation. A large second noise index means that the calibration spectrum contains less noise. A small second noise index means that the calibration spectrum contains more noise.

Next, in the second noise evaluation step, as shown by reference symbol S24 in FIG. 32, the second noise index is compared with the second noise threshold TH22. If the second noise index is equal to or greater than the second noise threshold TH22, the second processing unit 272 performs the second recording step. In the second recording step, as shown by reference symbol S25 in FIG. 32, a calibration spectrum having a second noise index equal to or greater than the second noise threshold TH22 is recorded.

The first generation unit 251 described above may generate the target spectrum by using the recorded calibration spectrum as a reference. This makes it possible to suppress the influence of noise in the calibration spectrum from appearing in the target spectrum.

If the second noise index is less than the second noise threshold TH22, it is considered that the entire or part of the surface of the calibration sample 51 is contaminated. In this case, as shown by the reference symbol S26 in FIG. 32, the calibration sample 51 may be changed to another one, or the measurement position on the calibration sample 51 may be changed. Then, steps S21, S22, S23, and S24 may be performed again. That is, the second irradiation step, the second detection step, the second generation step, and the second processing step may be performed again on a new calibration sample 51 or on a new measurement position of the calibration sample 51. According to the calibration step shown in FIG. 32, a calibration spectrum containing a lot of noise caused by contamination on the surface of the calibration sample 51 can be eliminated. In addition, the calibration step can be repeatedly performed until a calibration spectrum having a second noise index equal to or greater than the second noise threshold TH22 is obtained. Therefore, an appropriate calibration spectrum can be recorded.

The calibration process may be continued until N calibration spectra having a second noise index equal to or greater than the second noise threshold TH22 are obtained, as indicated by reference symbol S27 in FIG. 32. For example, as indicated by reference symbol S28 in FIG. 32, if the number of recorded calibration spectra is less than N, the measurement position on the calibration sample 51 may be changed. Then, steps S21, S22, S23, and S24 may be performed again. That is, the second irradiation step, the second detection step, the second generation step, and the second processing step may be performed again for the new measurement position on the calibration sample 51. According to the calibration process shown in FIG. 32, N calibration spectra having a second noise index equal to or greater than the second noise threshold TH22 can be obtained. N is, for example, 2 or more, may be 3 or more, or may be 4 or more. N may be, for example, 10 or less.

In the calibration process, as shown by reference symbol S28 in FIG. 32, the average of the N recorded calibration spectra may be calculated. This allows an averaged calibration spectrum to be generated. The spectral value at each wavelength point of the averaged calibration spectrum is the average of the spectral values of the N calibration spectra. For example, the value of the averaged calibration spectrum at a wavelength point of 1200 nm is the average of the values at the wavelength point of 1200 nm of the N calibration spectra.

The first generation unit 251 described above may generate the target spectrum by using the averaged calibration spectrum as a reference. This can further suppress the influence of noise in the calibration spectrum from appearing in the target spectrum.

(First Modification of the Fifth Embodiment)
The calibration process in the above-mentioned fifth embodiment may be combined with the second embodiment, that is, the noise contained in the calibration spectrum in the second embodiment may be evaluated by the above-mentioned calibration process.

(Second Modification of the Fifth Embodiment)
The calibration process in the above-mentioned fifth embodiment may be combined with the third embodiment, that is, the noise contained in the calibration spectrum in the third embodiment may be evaluated by the above-mentioned calibration process.

(Third Modification of the Fifth Embodiment)
The calibration process in the above-described fifth embodiment may be combined with the fourth embodiment, that is, the noise contained in the calibration spectrum in the fourth embodiment may be evaluated by the above-described calibration process.

(Example of the fifth embodiment)
(Example B1)
Using the above-mentioned inspection system 15, the calibration spectrum of the first calibration sample 51 was measured. Spectralon manufactured by Labsphere was used as the first calibration sample 51. A near-infrared spectroscopy built-in reflection module M-R2 manufactured by innoSpectra was used as the spectroscopic module 20 equipped with the irradiation device 21 and the detection device 24. The calibration process was carried out in a state where the calibration sample 51 was fixed to the case 40 by a rubber band 52 as shown in FIG.

The detector 24 detects the intensity of light at each wavelength point by decomposing the light into components corresponding to each wavelength point. The detector 24 scans each wavelength point in sequence. The conditions for the detection process are the same as those in Example A1.

Figure 33 shows the measurement results of the calibrated spectrum. The second noise index of the calibrated spectrum shown in Figure 33 was 38752. The second noise threshold TH22 was 25000. Therefore, it can be said that the calibrated spectrum shown in Figure 33 contains little noise. In the following description, the calibrated spectrum shown in Figure 33 is also referred to as spectrum 11.

Next, three more calibration spectra were measured with the measurement position of the first calibration sample 51 changed. The three calibration spectra are also referred to as spectrum 12, spectrum 13, and spectrum 14. The measurement position when spectrum 12 was measured was 2 mm away from the measurement position when spectrum 11 was measured. The measurement position when spectrum 13 was measured was 2 mm away from the measurement position when spectrum 12 was measured. The measurement position when spectrum 14 was measured was 2 mm away from the measurement position when spectrum 13 was measured. Spectra 11 to spectrum 14 are shown in Figure 34. The second noise indices of spectra 12, 13, and 14 were 38793, 35463, and 36466.

It can be said that the noise contained in spectra 11 to 14 is small. An averaged calibration spectrum may be generated by averaging spectra 11 to 14.

(Example B2)
The calibration spectrum of the second calibration sample 51, which is different from the first calibration sample 51, was measured using the above-described inspection system 15. The measurement conditions were the same as those in Example B1.

Figure 35 shows the measurement results of the calibrated spectrum. The second noise index of the calibrated spectrum shown in Figure 35 was 16724. Therefore, it can be said that the calibrated spectrum shown in Figure 35 contains a large amount of noise. In the following description, the calibrated spectrum shown in Figure 35 is also referred to as spectrum 21.

Next, three more calibration spectra were measured with the measurement position of the second calibration sample 51 changed. The three calibration spectra are also referred to as spectrum 22, spectrum 23, and spectrum 24. The measurement position when spectrum 22 was measured was 2 mm away from the measurement position when spectrum 21 was measured. The measurement position when spectrum 23 was measured was 2 mm away from the measurement position when spectrum 22 was measured. The measurement position when spectrum 24 was measured was 2 mm away from the measurement position when spectrum 23 was measured. Spectra 21 to spectrum 24 are shown in Figure 36. The second noise indices of spectra 22, 23, and 24 were 9198, 13066, and 17022, respectively.

(Example B3)
The above-described inspection system 15 was used to determine the composition of a first plurality of samples and a second plurality of samples, the first samples including a first candidate composition, and the second samples including a second candidate composition.

Target spectra including the absorbance of the multiple first samples and the multiple second samples were each generated at standard temperature. In the process of calculating the absorbance, an averaged calibration spectrum generated by the calibration process shown in FIG. 32 was used. The averaged calibration spectrum was generated by averaging five calibration spectra having a second noise index equal to or greater than the second noise threshold TH22.

Then, the first processing unit 271 performed a predetermined process on the target spectrum. As a result, parameter values were calculated at each wavelength point. Next, the analysis unit 282 applied a standard regression equation to the parameter values. As a result, second similarities were calculated for the multiple first samples and the multiple second samples, as shown in FIG. 37.

As shown in FIG. 37, the second similarity of the first sample containing the first candidate composition was smaller than the standard threshold value TH0. The second similarity of the second sample containing the second candidate composition was larger than the standard threshold value TH0. In other words, both the first sample and the second sample were appropriately determined.

(Comparative Example B1)
The second similarity of the plurality of first samples and the plurality of second samples was calculated by carrying out the same steps as in Example B3, except that the calibration spectra having the second noise index less than the second noise threshold TH22 were not excluded. The results are shown in FIG.

In the calibration process of the example shown in FIG. 38, in the measurement process of one of the five calibration spectra used to generate the averaged calibration spectrum, the reflectance spectrum of air was measured instead of the calibration sample 51. Therefore, the value of the averaged calibration spectrum is significantly different from that of Example B3. As a result, as shown in FIG. 38, the second similarities of the multiple first samples and the multiple second samples were all significantly greater than 1. In this case, the multiple first samples were erroneously determined to be the second candidate composition.

(Comparative Example B2)
The second similarity of the plurality of first samples and the plurality of second samples was calculated by carrying out the same steps as in Example B3, except that the calibration spectra having the second noise index less than the second noise threshold TH22 were not excluded. The results are shown in FIG.

In the calibration process of the example shown in FIG. 39, vibration was applied to the inspection system 15 when the reflectance spectrum was measured. For this reason, the value of the averaged calibration spectrum is significantly different from that in Example B3. As a result, as shown in FIG. 39, the second similarity of the multiple first samples and the multiple second samples both had values greater than 1. In this case, the multiple first samples were erroneously determined to be the second candidate composition.

Sixth embodiment
Fig. 40 is a perspective view showing an example of the case 40. As shown in Fig. 40, the case 40 may include a first friction layer 53 provided on the first surface 41 or the outer edge. In the example shown in Fig. 40, the first friction layer 53 is provided on the first surface 41 along the third outer edge 413 and the fourth outer edge 414.

The static friction force between the first friction layer 53 and the object 10 is greater than the static friction force between the first surface 41 and the object 10. By providing the first friction layer 53 on the case 40, it is possible to prevent the case 40 from moving during the inspection process.

The first friction layer 53 may have any configuration. For example, the first friction layer 53 may be formed of a surface layer of tape attached to the first surface 41. As the tape, an anti-slip tape, item no. 5187 manufactured by Mysto Co., Ltd., or the like may be used.

Seventh embodiment
Fig. 41 is a perspective view showing an example of the case 40. As shown in Fig. 41, the case 40 may include a support member 54 fixed to the side surface. In the example shown in Fig. 41, the case 40 includes one support member 54 fixed to the first side surface 43 and two support members 54 fixed to the second side surface 44. Although not shown, the case 40 may include one or two support members 54, or may include four or more support members 54. The support member 54 may be fixed to the third side surface 45 or the fourth side surface 46.

The support member 54 may include a first portion 541 fixed to the side surface, and a second portion 542 connected to the first portion 541. The first portion 541 may extend from the side surface toward the first surface 41. The second portion 542 may include a surface extending parallel to the first surface 41. The second portion 542 may contact the object 10 during the inspection process. This allows the position of the case 40 to be more stably maintained during the inspection process. This makes it possible to suppress noise caused by hand movement, etc. from appearing in the target spectrum.

As shown in FIG. 41, the case 40 may include a second friction layer 55 provided on the second portion 542. The second friction layer 55 includes, for example, rubber. The second friction layer 55 may contact the object 10 during the inspection process. The static friction force between the second friction layer 55 and the object 10 is greater than the static friction force between the first surface 41 and the object 10. By providing the second friction layer 55 on the second portion 542, movement of the case 40 during the inspection process can be further suppressed.

Eighth embodiment
Fig. 42 is a perspective view showing an example of the case 40. As shown in Fig. 42, the case 40 may include a cooling unit 56 that cools the irradiation device 21 and/or the detection device 24. The cooling unit 56 is, for example, a blower such as a fan. The cooling unit 29 may cool the irradiation device 21 and/or the detection device 24 by a Peltier element or a heat sink. By providing the cooling unit 56, it is possible to prevent the temperature inside the case 40 from becoming higher than the temperature of the surrounding atmosphere. This makes it possible to prevent the characteristics of the spectroscopy module 20 and the like from changing due to temperature.

When the ambient temperature in which the inspection system 15 is placed is low, there is little risk that the temperature of the spectroscopic module 20 will rise to a level that affects the intensity of the reflected light detected by the spectroscopic module 20. In this case, the irradiation device 21 and/or the detection device 24 do not need to be cooled by the cooling unit 56.

In addition, if the ambient temperature in which the inspection system 15 is placed is low and affects the intensity of the reflected light detected by the spectroscopic module 20, the irradiation device 21 and/or the detection device 24 may be covered with a highly insulating material to prevent a drop in temperature of the irradiation device 21 and/or the detection device 24. In other words, the inspection system 15 may be provided with an insulating material that covers the irradiation device 21 and/or the detection device 24. Alternatively, in this case, the irradiation device 21 and/or the detection device 24 may be heated by a heater or the like. In other words, the inspection system 15 may be provided with a heating unit that heats the irradiation device 21 and/or the detection device 24.

Furthermore, it is preferable that the spectroscopic module 20 detects the intensity of the reflected light after a certain amount of time has elapsed since the spectroscopic module 20 was powered on. In other words, it is preferable to warm up the spectroscopic module 20 before detecting the reflected light. In this case, the temperature of the irradiation device 21 and/or the detection device 24 becomes stable during the warm-up operation, and the intensity of the reflected light can be detected stably.

Ninth embodiment
In the above-described embodiment, an example has been shown in which the detection device 24 detects the intensity of light reflected by the object 10. However, this is not limiting, and although not shown, the detection device 24 may detect the intensity of light transmitted through the object 10. In this case, the inspection system 15 generates a spectrum including information on the intensity of light transmitted through the object 10. The inspection system 15 can determine the composition of the object 10 by performing the above-described inspection method based on the spectrum of the transmitted light. In addition, the inspection system 15 can be manufactured by performing the above-described manufacturing method based on the spectrum of the transmitted light.

The data at each wavelength point of the spectrum includes information regarding the intensity of transmitted light at the wavelength point. The data may be the intensity of transmitted light itself. The data may be calculated by performing some processing on the intensity of transmitted light at the wavelength point. For example, the data may be transmittance or absorbance Abs. The transmittance is a dimensionless quantity that represents the degree to which the object 10 transmits light. The transmittance T is calculated, for example, by the following formula.
T = I/I ₀
I is the intensity of light transmitted through the object 10 and incident on the detection device. _I0 is the intensity of light transmitted through a calibration sample and incident on the detection device. The calibration sample includes, for example, the base sheet 11 of the object 10, but does not include the resin layer 12. In this case, _I0 reflects the transmittance of air and the transmittance of the base sheet 11.
The absorbance Abs is calculated, for example, by the following formula.
Abs=-log ₁₀ (I/I ₀ )

The multiple components disclosed in the above embodiment and modified examples may be combined as necessary. Alternatively, some components may be deleted from all the components shown in the above embodiment and modified examples.

10 Object 10x First surface 10y Second surface 11 Base material sheet 12 Resin layer 15 Inspection system 20 Spectroscopic module 21 Irradiation device 211 Irradiation section 23 Temperature monitor 24 Detection device 241 Detection section 25 Generation device 251 First generation section 252 Second generation section 27 Processing device 271 First processing section 272 Second processing section 28 Determination device 281 Memory section 282 Analysis section 283 Determination section 40 Case 41 First surface 42 Second surface 48 Window 51 Calibration sample 53 First friction layer 54 Support member 541 First portion 542 Second portion 55 Second friction layer 56 Cooling section

Claims (13)

An inspection system for determining a composition of a resin layer of a resin sheet, comprising:
a temperature monitor for monitoring an ambient temperature;
An irradiation device including an irradiation unit that irradiates the resin sheet with near-infrared light as irradiation light;
a detection device that detects the intensity of light reflected by the resin sheet or light transmitted through the resin sheet;
a generation device for generating a spectrum containing information about the intensities at a plurality of wavelength points;
a processing device that calculates, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
a determination device including: a storage unit in which a standard regression equation expressing a relationship between a plurality of parameter values and a first candidate composition and a second candidate composition is stored; an analysis unit that calculates a similarity by applying the standard regression equation to the plurality of parameter values; and a determination unit that determines whether a composition of the resin layer is the first candidate composition or the second candidate composition based on a comparison between the similarity and a threshold value corresponding to the ambient temperature;
the second candidate composition corresponds to the twenty-first candidate composition and the twenty-second candidate composition;
the storage unit stores a sub-standard regression equation expressing a relationship between a parameter value at a standard temperature and the 21st candidate composition and the 22nd candidate composition;
When the composition of the resin layer is determined to be the second candidate composition, the analysis unit applies the sub-standard regression equation to the multiple parameter values to calculate a sub-similarity;
An inspection system in which the judgment unit determines whether the composition of the resin layer is the 21st candidate composition or the 22nd candidate composition based on a comparison between the sub-similarity and a sub-threshold corresponding to the ambient temperature. An inspection system for determining a composition of a resin layer of a resin sheet, comprising:
a temperature monitor for monitoring an ambient temperature;
An irradiation device including an irradiation unit that irradiates the resin sheet with near-infrared light as irradiation light;
a detection device that detects the intensity of light reflected by the resin sheet or light transmitted through the resin sheet;
a generation device for generating a spectrum containing information about the intensities at a plurality of wavelength points;
a processing device that calculates, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
a determination device including: a storage unit in which a plurality of regression equations are stored that express relationships at a plurality of temperatures between a parameter value and a first candidate composition and a second candidate composition; an analysis unit that selects a regression equation corresponding to the ambient temperature from the plurality of regression equations and applies the selected regression equation to the plurality of parameter values to calculate a similarity; and a determination unit that determines whether a composition of the resin layer is the first candidate composition or the second candidate composition based on the similarity;
the second candidate composition corresponds to the twenty-first candidate composition and the twenty-second candidate composition;
the storage unit stores a plurality of sub-regression equations that represent relationships between parameter values and the 21st candidate composition and the 22nd candidate composition at a plurality of temperatures;
When the composition of the resin layer is determined to be the second candidate composition, the analysis unit selects a sub-regression equation corresponding to the ambient temperature from the plurality of sub-regression equations, and applies the selected sub-regression equation to the plurality of parameter values to calculate a sub-similarity;
An inspection system, wherein the judgment unit judges whether the composition of the resin layer is the 21st candidate composition or the 22nd candidate composition based on the sub-similarity. The inspection system according to claim 1 , further comprising a cooling unit that cools the irradiation device and/or the detection device. The inspection system according to claim 1 , wherein the processing device includes 69 or more parameter values of the resin layer within a wavelength range of 800 nm to 2500 nm. The inspection system according to claim 1 , wherein the illumination device further includes a diffusion section that diffuses the illumination light. The inspection system of claim 1 , wherein the detection device includes a near-infrared spectrometer. The inspection system of claim 1 , wherein the detection device comprises a near-infrared hyperspectral camera. An inspection method for determining a composition of a resin layer of a resin sheet, comprising:
A first irradiation step of irradiating the resin sheet with near-infrared light as irradiation light;
a first detection step of detecting an intensity of light reflected by the resin sheet or light transmitted through the resin sheet;
a first generation step of generating a spectrum including information about the intensities at a plurality of wavelength points;
a first processing step of calculating, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained at a plurality of wavelength points by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
an analysis step of calculating a similarity by applying a standard regression equation that represents a relationship between a parameter value at a standard temperature and a first candidate composition and a second candidate composition to the plurality of parameter values;
and a determining step of determining whether the composition of the resin layer is the first candidate composition or the second candidate composition based on a comparison between the similarity and a threshold value corresponding to an ambient temperature,
the second candidate composition corresponds to the twenty-first candidate composition and the twenty-second candidate composition;
When the composition of the resin layer is determined to be the second candidate composition, the analysis step includes a step of calculating sub-similarity by applying a sub-standard regression equation that represents a relationship between a parameter value at a standard temperature and the 21st candidate composition and the 22nd candidate composition to the plurality of parameter values;
An inspection method, wherein the determination process includes a process of determining whether the composition of the resin layer is the 21st candidate composition or the 22nd candidate composition based on a comparison of the sub-similarity with a sub-threshold corresponding to the ambient temperature. An inspection method for determining a composition of a resin layer of a resin sheet, comprising:
A first irradiation step of irradiating the resin sheet with near-infrared light as irradiation light;
a first detection step of detecting an intensity of light reflected by the resin sheet or light transmitted through the resin sheet;
a first generation step of generating a spectrum including information about the intensities at a plurality of wavelength points;
a first processing step of calculating, as a plurality of parameter values corresponding to each wavelength point, values of the light intensity at a plurality of wavelength points or values obtained at a plurality of wavelength points by performing processing including averaging, smoothing, normalization, differentiation, scatter correction, baseline correction, peak shift correction, or a combination thereof on the spectrum;
an analysis step of selecting a regression equation corresponding to an ambient temperature from among a plurality of regression equations that express relationships between a parameter value and a first candidate composition and a second candidate composition at a plurality of temperatures, and applying the selected regression equation to the plurality of parameter values to calculate a similarity;
and determining whether the composition of the resin layer is the first candidate composition or the second candidate composition based on the similarity,
the second candidate composition corresponds to the twenty-first candidate composition and the twenty-second candidate composition;
When the composition of the resin layer is determined to be the second candidate composition, the analysis step includes a step of selecting a sub-regression equation corresponding to the ambient temperature from among a plurality of sub-regression equations expressing relationships between a parameter value and the 21st candidate composition and the 22nd candidate composition at a plurality of temperatures, and applying the selected sub-regression equation to the plurality of parameter values to calculate sub-similarity;
An inspection method, wherein the determining step includes a step of determining whether the composition of the resin layer is the 21st candidate composition or the 22nd candidate composition based on the sub-similarity. 10. The inspection method according to claim 8, wherein the first processing step calculates 69 or more parameter values of the resin layer within a wavelength range of 800 nm to 2500 nm. The inspection method according to claim 8 , wherein the first irradiation step includes diffusing the irradiation light. The inspection method according to claim 8 , wherein the first detection step acquires the spectrum using a near-infrared spectrometer. The inspection method according to claim 8 , wherein the first detection step acquires the spectrum using a near-infrared hyperspectral camera.