JP7706744B2 - Coating thickness measurement method using images, coating thickness measurement program, and coating thickness measurement system - Google Patents

Description

The present invention relates to a coating thickness measurement method using images, a coating thickness measurement program, and a coating thickness measurement system for measuring coating thickness using images.

In shipbuilding painting, it is particularly important that the specified coating thickness is applied reliably. Therefore, SI paints with a "Self-indicating (SI) function" have been developed as a way for workers at the painting site to visually check whether the specified coating thickness has been achieved. SI paints are paints in which the color of the coating changes stepwise until the specified coating thickness is reached by adjusting the pigment components in the paint, and it is said that the specified coating thickness can be reliably ensured by applying paint over and over until the color of the base is no longer visible.
Additionally, on-site coating thickness measurements have been performed using electromagnetic or ultrasonic coating thickness gauges (for dry coatings) or wet gauges (for wet coatings).

Patent Document 1 discloses a paint film thickness measuring device in which a surface roughness determination unit determines the surface roughness (a value corresponding to the film thickness), and a first film thickness calculation means normally calculates the paint film thickness based on the roughness, the amount of change in roughness over time, the wavelength distribution, and the painting conditions, but when the surface roughness is equal to or greater than a standard roughness (corresponding to a thin film thickness of 40 μm or less), a second film thickness calculation means calculates the paint film thickness based on the wavelength distribution.
Furthermore, Patent Document 2 discloses a coating thickness measuring device that measures the coating thickness of a spot on a coated surface in a non-contact manner, the device comprising a light source that irradiates light onto a coated surface formed by applying paint to a steel material to be coated, a measuring instrument that measures color information of the reflected light, and signal processing means that converts the color information obtained by the measuring instrument into coating thickness.
Patent Document 3 discloses a method for measuring the thickness of a coating film, in which a calibration curve of thickness is prepared in advance from the reflectance of ultraviolet light irradiated onto the surface of a standard test piece having a transparent coating film containing an ultraviolet absorber formed thereon, and the thickness of the coating film, and ultraviolet light is irradiated onto the surface of a test piece produced by applying a transparent paint having the same composition as the standard test piece to the surface of a substrate, and the thickness of the test piece is calculated based on the calibration curve from the reflectance of the ultraviolet light.
Patent Document 4 discloses a film measuring device in which copper sheets are successively unwound from a roll, coated with titanium oxide paint in a coating section, and dried in a drying oven, and then the film thickness of the titanium oxide is measured with a film measuring device, using a color CCD sensor to capture an image over the entire width direction of the sheet, and the resulting video signal is converted into gradation data for each of the RGB color components on a video board of a controller, and of these, the G or B data that is effective for measurement is compared by an arithmetic circuit with a reference value stored in a reference thickness table to determine the corresponding film thickness, and if it is outside the allowable range, a marker is used to mark it, thereby accurately measuring the film thickness and weight of the titanium oxide, and FB (feedback) controlling the coating amount in the coating section to maintain a constant film thickness.

Japanese Patent Application Publication No. 9-105612 JP 2001-116520 A JP 2003-4419 A JP 2007-61780 A

Coating thickness measurements using electromagnetic/ultrasonic coating thickness gauges and wet gauges are essentially point measurements, and it is difficult to evaluate the coating thickness over an entire surface (measure the distribution of coating thickness).
In addition, since the coating thickness measuring device of Patent Document 1 predicts the dry film thickness after drying by calculation from the measured wet film thickness, the numerical accuracy of the obtained dry film thickness may be low.
Furthermore, the coating thickness measuring device of Patent Document 2 merely measures the thickness based on the correlation between color information such as color intensity and the thickness of the coating, and therefore the measurement accuracy of the thickness of the coating may be insufficient.
Moreover, the film thickness measurement method of Patent Document 3 is intended for transparent coating films and cannot be applied to opaque coating films.
Furthermore, the film measuring device in Patent Document 4 is intended for the film thickness of film- or sheet-like substrates produced in factories, and is not applicable to shipbuilding painting, etc.
Therefore, an object of the present invention is to provide a coating thickness measurement method, a coating thickness measurement program, and a coating thickness measurement system using images, which are capable of performing non-contact, planar coating thickness measurement with high accuracy.

The coating thickness measurement method using an image according to claim 1 is a method for measuring coating thickness using an image, characterized in that it comprises a parameter determination step of irradiating light onto a plurality of test pieces having coating films coated with different thicknesses of the target paint, estimating and determining the scattering coefficient (S) and absorption coefficient (K) of the coating film, a coating thickness estimation equation derivation step of applying the scattering coefficient (S) and absorption coefficient (K) to an equation relating the reflected light intensity (R(T)) of the coating film and the coating thickness (T) based on the Kubelka-Munk theory to derive a coating thickness estimation equation, an image acquisition step of illuminating the coating film coated on the object to be measured and acquiring the reflected light as an image with an imaging means, a coating thickness calculation step of determining the reflected light intensity (R(T)) from the image and applying it to the coating thickness estimation equation to determine the coating thickness (T), and a coating thickness providing step of providing the determined coating thickness (T).
According to the present invention as defined in claim 1, non-contact and planar film thickness measurement can be performed with high accuracy using an image.

The present invention described in claim 2 is characterized in that the scattering coefficient (S) and the absorption coefficient (K) in the parameter determination step are estimated by preparing a sufficient number of test pieces for the expected film thickness, and estimating the scattering coefficient (S) and the absorption coefficient (K) using a nonlinear least squares method including the Levenberg-Marquard method based on the results of measuring the reflected light intensity (R(T)) and the coating film thickness (T).
According to the present invention as defined in claim 2, the scattering coefficient (S) and absorption coefficient (K) of a coating film can be estimated with high accuracy.

The present invention according to claim 3 is characterized in that the step of deriving the film thickness estimation equation further comprises a fitting step of fitting the film thickness estimation equation using a film thickness of a coating film that is actually measured.
According to the third aspect of the present invention, the coating thickness (T) can be calculated with even greater accuracy from a coating thickness estimation equation fitted using the measured coating thickness.

The present invention as set forth in claim 4 is characterized in that the light in the parameter determination step and the illumination in the image acquisition step are of the same condition.
According to the fourth aspect of the present invention, film thickness measurement can be performed under the same conditions without correcting the scattering coefficient (S) and the absorption coefficient (K) determined in the parameter determination step.

The present invention as defined in claim 5 is characterized in that the imaging means is a spectral camera.
According to the fifth aspect of the present invention, a spectral image of a specific wavelength band can be obtained, and the reflected light intensity (R(T)) for each wavelength band can be measured with high accuracy.

The present invention as described in claim 6 is characterized by having a sample image acquisition step of producing a plurality of coating film samples with a known thickness after coating with paint and acquiring sample images of the coating film samples acquired under the same conditions, and a parameter correction step of acquiring the plurality of coating film samples acquired in the image acquisition step as sample images at the time of measurement, comparing the sample images with the sample images at the time of measurement, and correcting the scattering coefficient (S) and absorption coefficient (K).
According to the present invention described in claim 6, by comparing a sample image of a coating sample with a sample image at the time of measurement and correcting the scattering coefficient (S) and absorption coefficient (K), it is possible to accurately measure the coating thickness (T) taking into account differences in measurement conditions between the sample image and the sample image at the time of measurement.

The seventh aspect of the present invention is characterized in that the thickness of the coating sample is approximately the upper limit and the lower limit of the thickness of the coating of the object to be measured.
According to the seventh aspect of the present invention, even if the number of coating film samples is small, the coating film thickness (T) can be measured accurately and with high precision.

The eighth aspect of the present invention is characterized in that saturation is used as the reflected light intensity (R(T)).
According to the eighth aspect of the present invention, even in the visible light region, the coating thickness (T) can be measured with higher accuracy based on the saturation as the reflected light intensity (R(T)).

The present invention as described in claim 9 is characterized in that the target paint is a paint having a wavelength band in which the reflected light intensity (R(T)) of the paint film does not become saturated within the range of film thickness of the paint film of the object to be measured.
According to the present invention as set forth in claim 9, by using a paint having a wavelength band that does not cause saturation of the reflected light intensity (R(T)), it is possible to measure film thickness even in a thick film region by selecting the wavelength band.

The coating thickness measurement program according to claim 10 is a program for measuring coating thickness from an image, characterized in that the program causes a computer to execute the following steps: a parameter acquisition step of irradiating light onto a plurality of test pieces having coating films of different thicknesses of the target paint, estimating the scattering coefficient (S) and absorption coefficient (K) of the coating film, and acquiring determined parameters; a coating thickness estimation equation derivation step of applying the scattering coefficient (S) and absorption coefficient (K) to a relational equation between the reflected light intensity (R(T)) of the coating film and the coating thickness (T) based on the Kubelka-Munk theory to derive a coating thickness estimation equation; an image acquisition step of illuminating the coating film applied to the object to be measured, and acquiring the reflected light as an image with an imaging means; a coating thickness calculation step of determining the reflected light intensity (R(T)) from the image, applying the determined reflected light intensity (R(T)) to the coating thickness estimation equation to determine the coating thickness (T); and a coating thickness providing step of providing the determined coating thickness (T) .
According to the tenth aspect of the present invention, the coating thickness measuring method using an image can be executed more accurately and quickly, and the coating thickness can be measured in a non-contact manner with high accuracy.

The present invention according to claim 11 provides a method for producing a plurality of coating film samples having a known thickness by applying the paint to a computer, and acquiring sample images of the coating film samples acquired under the same conditions;
The method is characterized in that the multiple coating film samples acquired in the image acquisition step are also acquired as sample images at the time of measurement, and a parameter correction step is executed in which the sample images and the sample images at the time of measurement are compared to correct the scattering coefficient (S) and the absorption coefficient (K) .
According to the present invention described in claim 11 , by correcting the scattering coefficient (S) and the absorption coefficient (K), the coating thickness (T) can be measured accurately and quickly, taking into account the difference in measurement conditions between the sample image and the sample image at the time of measurement.

The coating thickness measurement system using images according to claim 12 is a system for measuring coating thickness using images, characterized in that it comprises: a coating thickness estimation equation derivation means for applying the scattering coefficient (S) and absorption coefficient (K) of the coating of a target paint to an equation relating the reflected light intensity (R(T)) of the coating and the coating thickness (T) based on the Kubelka-Munk theory to derive a coating thickness estimation equation; an image acquisition means for illuminating the coating of the object to be measured and acquiring the reflected light as an image with an imaging means; a coating thickness calculation means for determining the reflected light intensity (R(T)) from the image and applying the resulting reflected light intensity to the coating thickness estimation equation to determine the coating thickness (T); and a coating thickness providing means for providing the determined coating thickness (T).
According to the twelfth aspect of the present invention , non-contact and planar film thickness measurement can be performed with high accuracy using an image.

The present invention according to claim 13 is characterized in that the illumination in the image acquiring means is light having the same conditions as those when the scattering coefficient (S) and the absorption coefficient (K) are calculated.
According to the thirteenth aspect of the present invention , film thickness measurement can be performed under the same conditions without correcting the determined scattering coefficient (S) and absorption coefficient (K).

The present invention as defined in claim 14 is characterized in that the imaging means is a spectral camera.
According to the fourteenth aspect of the present invention, a spectral image of a specific wavelength band can be obtained, and the reflected light intensity (R(T)) for each wavelength band can be measured with high accuracy.

The present invention as set forth in claim 15 is characterized in that it comprises a sample image storage means for storing sample images of a plurality of coating film samples having known film thicknesses, the sample images being acquired under identical conditions, and a parameter correction means for comparing the sample images acquired by the image acquisition means during measurement with the sample images, and correcting the scattering coefficient (S) and the absorption coefficient (K).
According to the present invention described in claim 15 , by comparing a sample image of a coating sample with a sample image at the time of measurement and correcting the scattering coefficient (S) and absorption coefficient (K), it is possible to accurately measure the coating thickness (T) taking into account differences in measurement conditions between the sample image and the sample image at the time of measurement.

The sixteenth aspect of the present invention is characterized in that saturation is used as the reflected light intensity (R(T)).
According to the sixteenth aspect of the present invention, the coating thickness (T) can be measured with higher accuracy based on the saturation as the reflected light intensity (R(T)) even in the visible light region.

The coating thickness measurement method using images of the present invention allows for highly accurate non-contact, planar coating thickness measurement using images.

In addition, when estimating the scattering coefficient (S) and absorption coefficient (K) in the parameter determination step, if a sufficient number of test pieces are prepared for the expected film thickness and the estimation is performed using a nonlinear least squares method including the Levenberg-Marquard method based on the results of measuring the reflected light intensity (R(T)) and the coating thickness (T), the scattering coefficient (S) and absorption coefficient (K) of the coating can be estimated with high accuracy.

In addition, if the film thickness estimation equation derivation step further includes a fitting step of fitting the film thickness estimation equation using the actually measured film thickness of the coating, the coating thickness (T) can be calculated with even greater accuracy from the film thickness estimation equation fitted using the measured film thickness.

In addition, if the light in the parameter determination step and the illumination in the image acquisition step are of the same conditions, film thickness measurement can be performed under the same conditions without correcting the scattering coefficient (S) and absorption coefficient (K) determined in the parameter determination step.

In addition, if the imaging means is a spectral camera, it is possible to obtain a spectral image of a specific wavelength band and accurately measure the reflected light intensity (R(T)) for each wavelength band.

In addition, if there is a sample image acquisition step in which multiple coating film samples with known coating thicknesses are produced and sample images of the coating film samples acquired under the same conditions are acquired, and the multiple coating film samples acquired in the image acquisition step are also acquired as sample images at the time of measurement, and the sample images are compared with the sample images at the time of measurement to correct the scattering coefficient (S) and the absorption coefficient (K), then the sample images of the coating film samples are compared with the sample images at the time of measurement to correct the scattering coefficient (S) and the absorption coefficient (K), thereby enabling the coating film thickness (T) to be measured with high accuracy, taking into account differences in measurement conditions between the sample images and the sample images at the time of measurement.

In addition, when the thickness of the coating sample is approximately the upper and lower limits of the coating thickness of the object being measured, the coating thickness (T) can be measured accurately and with higher precision even with a small number of coating samples.

In addition, when saturation is used as the reflected light intensity (R(T)), the coating thickness (T) can be measured with higher accuracy based on the saturation as the reflected light intensity (R(T)), even in the visible light range.

In addition, if the paint being measured has a wavelength band in which the reflected light intensity (R(T)) of the paint is not saturated within the range of the paint thickness of the object being measured, then by using a paint having a wavelength band in which the reflected light intensity (R(T)) is not saturated, and by selecting the wavelength band, it is possible to measure the paint thickness even in thick film regions.

In addition, the coating thickness measurement program using images of the present invention allows the coating thickness measurement method using images to be executed more accurately and quickly, enabling non-contact and highly accurate coating thickness measurement.

In addition, when the film thickness estimation equation deriving step is performed by a computer, the film thickness estimation equation fitted using the measured film thickness can be derived accurately and quickly.

In addition, when the computer is caused to execute the sample image acquisition step and the parameter correction step, the scattering coefficient (S) and absorption coefficient (K) are corrected, so that the coating thickness (T) can be measured quickly and accurately, taking into account the difference in measurement conditions between the sample image and the sample image at the time of measurement.

In addition, the image-based coating thickness measurement system of the present invention can perform non-contact, planar coating thickness measurement with high accuracy using images.

In addition, if the illumination in the image acquisition means is the same as that used when calculating the scattering coefficient (S) and absorption coefficient (K), film thickness measurement can be performed under the same conditions without correcting the calculated scattering coefficient (S) and absorption coefficient (K).

In addition, if the imaging means is a spectral camera, it is possible to obtain a spectral image of a specific wavelength band and accurately measure the reflected light intensity (R(T)) for each wavelength band.

In addition, when a sample image storage means is provided for storing sample images of multiple coating film samples with known film thicknesses acquired under the same conditions, and a parameter correction means is provided for comparing the sample images at the time of measurement of multiple coating film samples acquired by the image acquisition means during measurement with the sample images, and correcting the scattering coefficient (S) and absorption coefficient (K), the coating film thickness (T) can be accurately measured by comparing the sample images of the coating film samples with the sample images at the time of measurement and correcting the scattering coefficient (S) and absorption coefficient (K) while taking into account the difference in measurement conditions between the sample images and the sample images at the time of measurement.

In addition, when saturation is used as the reflected light intensity (R(T)), the coating thickness (T) can be measured with higher accuracy based on the saturation as the reflected light intensity (R(T)), even in the visible light range.

1 is a flow diagram of a first coating thickness measurement method using an image according to an embodiment of the present invention; Functional block diagram showing the means by which the system for measuring coating thickness using images is implemented A graph showing the predicted values of the K-M theory based on the values determined by parameter estimation for the color change (chroma) of the same paint. A graph showing the film thickness vs. reflected light intensity when the scattering coefficient (S) and absorption coefficient (K) are changed. Flow diagram of the second coating thickness measurement method using the same image

A coating thickness measuring method using an image, a coating thickness measuring program, and a coating thickness measuring system according to an embodiment of the present invention will be described.
FIG. 1 is a flow diagram of a first coating thickness measurement method using an image, and FIG. 2 is a functional block diagram showing a coating thickness measurement system using an image in terms of function realization means.
The coating thickness measurement system shown in Figure 2 is a system that measures coating thickness using images, and is equipped with a coating thickness estimation equation derivation means 10, an image acquisition means 20, a coating thickness calculation means 30, a coating thickness providing means 40, a sample image storage means 50, a parameter correction means 60, an imaging means 70, an input means 80, and a display means 90, and executes a part of the coating thickness measurement method using images.
The computer 100 is provided with the functions of an image acquisition means 20, a coating thickness calculation means 30, a coating thickness providing means 40, a sample image storage means 50, and a parameter correction means 60, and is installed with a coating thickness measurement program using images.
The film thickness estimation equation derivation means 10 applies the scattering coefficient (S) and absorption coefficient (K) of the coating of the paint to be measured to a relational equation between the reflected light intensity (R(T)) of the coating and the coating thickness (T) based on the Kubelka-Munk theory, to derive a film thickness estimation equation.
The image acquisition means 20 illuminates the coating film 111 of the object 110 to be measured, and the imaging means 70 acquires the reflected light as an image.
The coating thickness calculation means 30 obtains the reflected light intensity (R(T)) from the acquired image, and applies it to a coating thickness estimation formula to obtain the coating thickness (T).
The coating thickness providing means 40 provides the determined coating thickness (T) to the display means 90 .

As shown in Fig. 1, the first coating thickness measurement method first irradiates light onto a plurality of test pieces 1 having coating films of the target paint with different thicknesses, and estimates and determines the scattering coefficient (S) and absorption coefficient (K) of the coating film (S1: parameter determination step). The plurality of test pieces 1 are coated with the paint to be measured with different thicknesses, and the coating thicknesses are known.
The scattering coefficient (S) and the absorption coefficient (K) are parameters that are assumed to take different values for each paint.
The scattering coefficient (S) is the rate of increase in reflectance relative to an increase in coating thickness (T). As an extreme example, in the case of a transparent coating, the scattering coefficient (S) = 0 (no light reflection within the coating). For example, the intensity of reflected light from pigment particles in paint can be evaluated by the scattering coefficient (S). Even with the same pigment particles, the scattering coefficient (S) may increase as the content increases, and it is also thought that the scattering coefficient (S) will increase if highly reflective pigment particles are used.
The absorption coefficient (K) is the rate of decrease in transmittance relative to an increase in coating thickness (T), and is like the attenuation of light incident on the coating film 111. Since each resin has a different absorption spectrum, the value of the absorption coefficient (K) differs for each type of resin, and it is also assumed that the manner of attenuation differs depending on the wavelength band of the incident light.

The parameter determination step S1 can also be performed in a laboratory experiment, and the scattering coefficient (S) and absorption coefficient (K) of the coating film can be obtained in advance in a laboratory experiment before measuring the coating thickness on site. If the scattering coefficient (S) and absorption coefficient (K) are measured in advance for the combination of the coating material and light source to be measured, it becomes possible to measure the coating thickness distribution by one-shot photography on site by adjusting the photography conditions, such as setting the various conditions of the photography equipment and the light source 120 to the same as those in the laboratory experiment.
The processing procedure in the parameter determination step S1 is, for example, as follows.
1) Prepare a sufficient number of test pieces 1 for the film thickness range to be measured, irradiate each of the prepared test pieces 1 with light from a light source, and photograph the test pieces 1 using a spectral camera or the like.
2) Based on the captured image, data on "film thickness - reflected light intensity" is created.
3) Using the created "film thickness - reflected light intensity" data, parameters are estimated using a nonlinear least squares method such as the Levenberg-Marquard method to determine the scattering coefficient (S) and absorption coefficient (K) of the coating film.
In this manner, in the parameter determination step S1, the scattering coefficient (S) and absorption coefficient (K) of the coating film can be estimated with high accuracy by preparing a sufficient number of test pieces 1 for the expected film thickness and estimating the scattering coefficient (S) and absorption coefficient (K) using a nonlinear least squares method including the Levenberg-Marquard method based on the measurement results.

After the parameter determination step S1, the determined scattering coefficient (S) and absorption coefficient (K) are input to the computer 100 using the input means 80. The input means 80 is, for example, a keyboard or a mouse. The computer 100 applies the scattering coefficient (S) and absorption coefficient (K) to a relational equation between the reflected light intensity (R(T)) of the coating film and the coating film thickness (T) based on the Kubelka-Munk theory, and derives a film thickness estimation equation (S2: film thickness estimation equation derivation step).
In the film thickness estimation equation derivation step S2, the film thickness estimation equation derivation means 10 models the relationship between film thickness and reflected light intensity using the Kubelka-Munk theory. The Kubelka-Munk theory is a theory that relates the intensity of light incident on a material to the intensity of the light reflected from it, and is widely used in color matching of paints, etc. In the Kubelka-Munk theory, the relationship between film thickness (T) and reflected light intensity (R(T)) is expressed by the following equation (1).
The inverse function of formula (1) can be written as formula (2) below, which is the film thickness estimation formula, where ln is the natural logarithm.

Here, Fig. 3 shows the predicted values of the K-M theory based on the values determined by parameter estimation for the color change (saturation) of the paint, and shows the relationship between the coating film thickness (dry film thickness) and saturation. Note that saturation can also be called the reflected light intensity (R(T)) in the visible light region.
The "Measured data" in FIG. 3 is a value measured by a commercially available electromagnetic film thickness meter, and shows the average value and standard deviation of 30 measurements taken on one test piece 1. From FIG. 3, it can be seen that the color change of the coating film is well captured. As the film thickness increases, it approaches a constant value (corresponding to color saturation), and the estimation error increases in the region close to the value (R _∞ ) at which the reflected light intensity (R(T)) converges. The value (R _∞ ) at which the reflected light intensity (R(T)) converges depends on the values of the scattering coefficient (S) and the absorption coefficient (K). In this way, by using saturation as the reflected light intensity (R(T)), the coating film thickness (T) can be measured with high accuracy based on the saturation as the reflected light intensity (R(T)) even in the visible light region. That is, the coating film thickness (T) can be obtained based on the K-M theory even in the visible light region by utilizing saturation. Therefore, the coating film thickness (T) can be measured with high accuracy without using a spectral camera. In addition, the coating thickness (T) may be measured by combining the reflected light intensity (R(T)) obtained by a spectral camera or hyperspectral camera with the chroma obtained in the visible light band. If there are color samples of about two different thicknesses, it is possible to estimate the coating thickness from the image.
Furthermore, the film thickness estimation equation derivation step S2 includes a fitting step S2-1 in which the film thickness estimation equation is fitted using the actually measured film thickness of the coating film 111. "K-M fitting" in Fig. 3 is the result of estimating the parameters of the K-M theory (Equation (1)) using the L-M method and fitting it to the "Measured data". By fitting the film thickness estimation equation in this way, the coating thickness (T) can be calculated with even greater accuracy from Equation (2), which is the film thickness estimation equation fitted using the measured film thickness.
The scattering coefficient (S) and absorption coefficient (K) are expected to vary depending on the material of the coating film 111 (paint components such as pigments and resins) and the measurement conditions (photography equipment, light source 120, etc.), as well as the wavelength of the reflected light (the shape of the fitting function in Figure 3 changes for each wavelength).

SI paint (Self Indication paint) used for painting ballast tanks on ships is designed to saturate color change at a film thickness of 320 μm, which is required by the IMO (International Maritime Organization) Paint Performance Standard (PSPC). Therefore, in the film thickness range of 320 μm or more, the color does not change even if the film thickness increases, making it impossible to distinguish the film thickness with the naked eye.
However, depending on the wavelength of the reflected light, it is possible that a change different from the color change discernible by the naked eye may occur, for example, in a certain wavelength band, where the reflected light intensity (R(T)) does not saturate even in a range exceeding a film thickness of 320 μm. Therefore, by selecting a wavelength band in which the target paint has a wavelength band in which the reflected light intensity (R(T)) of the coating film 111 does not saturate in the range of film thickness of the coating film 111 of the object 110 to be measured, and by using a wavelength band in which the reflected light intensity (R(T)) does not saturate, it becomes possible to measure the film thickness in a thick film region (320 μm or more) where the film thickness cannot be discerned by the naked eye.
Here, FIG. 4 is a graph showing the "film thickness-reflected light intensity" when the values of the scattering coefficient (S) and the absorption coefficient (K) are changed. Line A shows the original values of the scattering coefficient (S) and the absorption coefficient (K) calculated in the parameter determination step S1, line B shows the case where the scattering coefficient (S) and the absorption coefficient (K) are each twice the original value, and line C shows the case where the scattering coefficient (S) and the absorption coefficient (K) are each 0.5 times the original value. In either case, the film converges to a constant value, but the smaller the scattering coefficient (S) and the absorption coefficient (K) are, the larger the film thickness value to which the film converges. In other words, by combining the material and wavelength band of the coating film 111 that reduces the scattering coefficient (S) and the absorption coefficient (K), it may be possible to measure the film thickness even in a range exceeding 320 μm in thickness.
The value (R _∞ ) at which the reflected light intensity (R(T)) converges is expressed by the following equation (3).
By slightly modifying equation (3), it can be seen that R _∞ can be expressed as a function of only (K/S). In other words, even if the values of the scattering coefficient (S) and the absorption coefficient (K) change, if the ratio between the scattering coefficient (S) and the absorption coefficient (K) is constant, R _∞ will be a constant value as shown in FIG.

After step S2 of deriving the film thickness estimation equation, a coating film 111 applied to an object 110 to be measured is illuminated by a light source 120, and the reflected light is captured as an image by an imaging means 70 (S3: image capturing step).
In the image acquisition step S3, the image acquisition means 20 images an area of, for example, about 1 cm square to 100 cm square of the object 110. The shooting distance is set to a predetermined range of, for example, about 1.0 m to 2.0 m, and the focal length is arbitrary.
Here, if the light in the parameter determination step S1 and the illumination in the image acquisition step S3 are of the same condition, the conditions of the light irradiated to the test piece 1 and the conditions of the light irradiated to the object 110 are the same, so that the film thickness measurement can be performed under the same conditions without correcting the scattering coefficient (S) and the absorption coefficient (K) determined in the parameter determination step S1. It is preferable to use a diffuse light source for photography, which has little unevenness in light intensity, for the light source used in the parameter determination step S1 and the light source 120 used in the image acquisition step S3. Also, the light source used in the parameter determination step S1 and the light source 120 used in the image acquisition step S3 may be the same light source.
The light irradiated in the parameter determination step S1 and the illumination in the image acquisition step S3 can be any light that can provide a reflected light intensity (R(T)) appropriate for determining the coating thickness (T), such as white light, light containing a specific wavelength, or light of a specific wavelength.

In this embodiment, the imaging means 70 used in the image acquisition step S3 is a spectral camera. This makes it possible to acquire a spectral image of a specific wavelength band and to accurately measure the reflected light intensity (R(T)) for several different wavelength bands.
In addition, as a simple method for measuring the reflected light intensity (R(T)) of several wavelength bands, the imaging means 70 can be a visible light camera, such as a general digital camera, equipped with a bandpass filter that transmits only light of a specific wavelength band.
As described above, the scattering coefficient (S) and the absorption coefficient (K) differ for each wavelength of the reflected light, and there may be a wavelength band in which the reflected light intensity (R(T)) is not saturated even in a range exceeding a film thickness of, for example, 320 μm. Therefore, it is preferable to use a multi-wavelength spectroscopic camera (hyperspectral camera) capable of measuring the spectrum of multiple wavelengths as the imaging means 70. This allows film thicknesses of, for example, 320 μm or more to be measured with even greater accuracy. In addition, by using a spectroscopic camera or hyperspectral camera with a large number of bands, it is possible that a color sample is not required. In addition, a light field camera capable of obtaining images with a different focus or stereoscopic images after shooting can be used alone or in combination with a spectroscopic camera or hyperspectral camera.

After the image acquisition step S3, the reflected light intensity (R(T)) is obtained from the image and applied to the film thickness estimation formula to obtain the coating thickness (T) (S4: coating thickness calculation step).
The parameters, scattering coefficient (S) and absorption coefficient (K), are determined in parameter determination step S1 and are therefore known. Therefore, by measuring the reflected light intensity (R(T)) from the surface of the coating film 111 with a light intensity sensor, the coating film thickness (T) can be calculated from equation (2) using coating film thickness calculation means 30.

After the coating thickness calculation step S4, the determined coating thickness (T) is provided from the coating thickness providing means 40 to the display means 90 (S5: coating thickness providing step).
The display means 90 is, for example, a printer or a monitor connected to the computer 100, and processes and outputs information regarding the coating thickness (T) provided by the coating thickness providing means 40.

Next, a coating thickness measurement method different from the flow shown in Fig. 1 will be described. Fig. 5 is a flow diagram of the second coating thickness measurement method using an image. Note that the same parts as those in the first coating thickness measurement method are given the same reference numerals and their explanations will be omitted.
The second coating thickness measurement method also utilizes a sample image storage means 50 and a parameter correction means 60 shown in Fig. 2. The sample image storage means 50 stores sample images 3 acquired under the same conditions of images of multiple coating film samples 2 with known film thicknesses. The parameter correction means 60 compares sample images 4 acquired by the image acquisition means 20 during measurement with the sample images 3, and corrects the scattering coefficient (S) and absorption coefficient (K).

When the conditions in the image acquisition step S3 (when photographing on-site) are different from the conditions in the parameter determination step S1 (when conducting a laboratory experiment), several coating film samples 2 with known film thicknesses are photographed simultaneously with the object 120 (subject to be photographed) (they are captured in the same image), and the scattering coefficient (S) and absorption coefficient (K) parameters can be corrected by referring to the reflected light intensity (R(T)) from the coating film samples 2. In other words, the difference in photographing conditions between the parameter determination step S1 and the image acquisition step S3 can be corrected.
Therefore, as shown in FIG. 5, in the second coating thickness measurement method, a sample image acquisition step S6 and a parameter correction step S7 are executed.
In the sample image acquisition step S6, a plurality of coating film samples 2 with known coating thicknesses are produced, and the coating film samples 2 are acquired under the same conditions as sample images 3 and stored in the sample image storage means 50.
In the parameter correction step S7, the parameter correction means 60 also acquires multiple coating film samples 2 as sample images 4 at the time of measurement in the image acquisition step S3, compares the sample images 3 with the sample images 4 at the time of measurement, and corrects the scattering coefficient (S) and absorption coefficient (K).
As a result, by comparing the sample image 3 of the coating sample with the sample image 4 at the time of measurement and correcting the scattering coefficient (S) and absorption coefficient (K), it is possible to accurately measure the film thickness distribution of the coating film (T) taking into account the difference in measurement conditions between the sample image 3 and the sample image 4 at the time of measurement. Note that, in order to correct the scattering coefficient (S) and absorption coefficient (K), automatic detection of each coating film sample 2 in the captured image and recognition and classification processing linked to the film thickness are required, so it is preferable to attach an identification marker such as a QR code (registered trademark) to each coating film sample 2.
It is also preferable to adopt a value for the film thickness value of the coating film sample 2 that will have a large difference within the film thickness range to be measured. Therefore, the film thickness of the coating film sample 2 is set to be approximately the upper limit and the lower limit of the film thickness of the coating film 111 of the object 110 to be measured. This allows the film thickness distribution of the coating film (T) to be measured accurately and with high precision even with a small number of coating film samples. For example, when the film thickness range to be measured is 0-450 μm, coating film samples 2 of 50 μm and 400 μm are prepared.

As described above, according to the coating thickness measurement method and coating thickness measurement system using images of the present invention, non-contact, planar coating thickness measurement can be performed with high accuracy using images.
Furthermore, by using an image-based coating thickness measurement program to have the computer 100 execute the image acquisition step S3, coating thickness calculation step S4, and coating thickness providing step S5 in the coating thickness measurement method, the image-based coating thickness measurement method can be executed more accurately and quickly, and coating thickness measurements can be performed non-contact and with high accuracy.
Furthermore, by causing the computer 100 to execute the film thickness estimation equation deriving step S2, it is possible to accurately and quickly derive a film thickness estimation equation that is fitted using the measured film thickness.
Furthermore, by having the computer 100 execute the sample image acquisition step S6 and the parameter correction step S7, the scattering coefficient (S) and the absorption coefficient (K) are corrected, and the coating thickness (T) can be measured accurately and quickly, taking into account the difference in measurement conditions between the sample image 3 and the sample image 4 at the time of measurement.

The present invention constructs an approximation function based on the Kubelka-Munk theory and derives a coating thickness estimation formula, making it possible to accurately estimate coating thickness for shipbuilding coatings and other coatings.

1 Test piece 2 Coating sample 3 Sample image 4 Sample image during measurement 10 Film thickness estimation formula deriving means 20 Image acquisition means 30 Coating thickness calculation means 40 Coating thickness providing means 50 Sample image storage means 60 Parameter correction means 70 Imaging means 100 Computer 110 Object 111 Coating S1 Parameter determination step S2 Film thickness estimation formula deriving step S2-1 Fitting step S3 Image acquisition step S4 Coating thickness calculation step S5 Coating thickness providing step S6 Sample image acquisition step S7 Parameter correction step

Claims (16)

A method for measuring coating thickness by an image, comprising the steps of:
A parameter determination step of irradiating light onto a plurality of test pieces having coating films of a target paint with different thicknesses to estimate and determine the scattering coefficient (S) and absorption coefficient (K) of the coating film;
a film thickness estimation equation deriving step of applying the scattering coefficient (S) and the absorption coefficient (K) to a relational equation between the reflected light intensity (R(T)) of the coating film and the coating film thickness (T) based on the Kubelka-Munk theory to derive a film thickness estimation equation;
an image acquisition step of illuminating the coating film applied to the object to be measured and acquiring an image of the reflected light by an imaging means;
A coating thickness calculation step of calculating a reflected light intensity (R(T)) from the image and applying it to the coating thickness estimation formula to calculate the coating thickness (T);
and providing the determined coating thickness (T). The method for measuring a coating thickness from an image according to claim 1, characterized in that the scattering coefficient (S) and the absorption coefficient (K) are estimated in the parameter determination step by preparing a sufficient number of test pieces for the expected coating thickness, and estimating the scattering coefficient (S) and the absorption coefficient (K) using a nonlinear least squares method including the Levenberg-Marquard method based on the results of measuring the reflected light intensity (R(T)) and the coating thickness (T). The method for measuring a coating thickness from an image according to claim 1 or 2, characterized in that the step of deriving the coating thickness estimation equation further includes a fitting step of fitting the coating thickness estimation equation using an actually measured coating thickness of the coating. The method for measuring a coating thickness using an image according to any one of claims 1 to 3, characterized in that the light in the parameter determination step and the illumination in the image acquisition step are light of the same conditions. The method for measuring a coating thickness using an image according to any one of claims 1 to 4, characterized in that the imaging means is a spectral camera. 6. The coating thickness measuring method using images according to claim 1, further comprising: a sample image acquiring step of acquiring sample images of the coating samples acquired under identical conditions, the sample images acquired in the image acquiring step being also acquired as sample images at the time of measurement, and a parameter correction step of comparing the sample images with the sample images at the time of measurement to correct the scattering coefficient (S) and the absorption coefficient (K). The method for measuring a coating thickness from an image according to claim 6, characterized in that the coating thickness of the coating sample is the coating thickness of the object to be measured at approximately the upper and lower limits. A coating thickness measurement method using an image according to any one of claims 1 to 7, characterized in that saturation is used as the reflected light intensity (R(T)). The method for measuring a coating thickness from an image according to any one of claims 1 to 8, characterized in that the coating to be measured is a coating having a wavelength band in which the reflected light intensity (R(T)) of the coating does not become saturated within the range of the coating thickness of the coating of the object to be measured. A program for measuring a coating thickness from an image,
On the computer,
a parameter acquisition step of irradiating light onto a plurality of test pieces having coating films of a target paint with different thicknesses, and estimating and acquiring parameters determined by estimating the scattering coefficient (S) and the absorption coefficient (K) of the coating film;
a film thickness estimation equation deriving step of applying the scattering coefficient (S) and the absorption coefficient (K) to a relational equation between the reflected light intensity (R(T)) of the coating film and the coating film thickness (T) based on the Kubelka-Munk theory to derive a film thickness estimation equation;
an image acquisition step of illuminating the coating film applied to the object to be measured and acquiring an image of the reflected light by an imaging means;
A coating thickness calculation step of calculating a reflected light intensity (R(T)) from the image and applying it to the coating thickness estimation formula to calculate the coating thickness (T);
and a coating thickness providing step of providing the determined coating thickness (T) . The computer includes:
a sample image acquisition step of producing a plurality of coating film samples having a known thickness by applying the paint and acquiring sample images of the coating film samples acquired under the same conditions;
11. The coating thickness measurement program according to claim 10, further comprising: acquiring a plurality of coating samples acquired in the image acquisition step as sample images at the time of measurement; and executing a parameter correction step of comparing the sample images with the sample images at the time of measurement to correct the scattering coefficient (S) and the absorption coefficient (K) . A system for measuring coating thickness using an image, comprising:
a film thickness estimation equation deriving means for deriving a film thickness estimation equation by applying the scattering coefficient (S) and absorption coefficient (K) of a coating film of a target paint to a relational equation between the reflected light intensity (R(T)) of the coating film and the coating film thickness (T) based on the Kubelka-Munk theory;
an image acquisition means for illuminating the coating film of the object to be measured and acquiring an image of the reflected light by an image acquisition means;
a coating thickness calculation means for calculating a reflected light intensity (R(T)) from the image and applying the calculated reflected light intensity (R(T)) to the coating thickness estimation formula to calculate the coating thickness (T);
and a coating thickness providing means for providing the determined coating thickness (T). 13. The coating thickness measurement system according to claim 12 , wherein the illumination in the image acquisition means is light having the same conditions as when the scattering coefficient (S) and the absorption coefficient (K) are calculated. 14. The coating thickness measuring system according to claim 12 , wherein the imaging means is a spectral camera. 15. The coating thickness measurement system using images according to claim 12, further comprising: a sample image storage means for storing sample images of a plurality of coating samples having known thicknesses, the sample images being acquired under identical conditions ; and a parameter correction means for comparing the sample images of the plurality of coating samples acquired by the image acquisition means during measurement with the sample images, and correcting the scattering coefficient (S) and the absorption coefficient (K). 16. The coating thickness measurement system according to claim 12 , wherein saturation is used as the reflected light intensity (R(T)).