JP7679706B2 - Visual Inspection Equipment - Google Patents

Description

The present invention relates to an appearance inspection device.

WO2019/216031A1 discloses an invention relating to a defect inspection device and a defect inspection method for ceramic bodies. The device disclosed here includes a table, an imaging unit, a low-angle illumination unit, a medium-angle illumination unit, and a high-angle illumination unit, a judgment image generation unit, and a defect judgment unit.

The table is a table on which the ceramic body to be inspected is placed. The imaging unit images at least a portion of the inspection surface of the ceramic body placed on the table as an imaging area from the normal direction of the inspection surface. The low-angle illumination unit, the medium-angle illumination unit, and the high-angle illumination unit each have four or more unit illuminations that each irradiate the imaging area with illumination light obliquely from different irradiation directions equiangularly spaced from one another around the imaging unit. The judgment image generation unit generates judgment image data for judging the presence or absence of defects in the imaging area based on the imaging data acquired by the imaging unit. The defect judgment unit judges the presence or absence of defects based on the judgment image data.

In the device disclosed herein, the multiple unit lights in each of the low-angle lighting unit, the medium-angle lighting unit, and the high-angle lighting unit are turned on and off in sequence. The imaging unit captures an image of the imaging area each time the multiple unit lights in each of the low-angle lighting unit, the medium-angle lighting unit, and the high-angle lighting unit are turned on, thereby generating multiple pieces of imaging data with low-angle lighting, multiple pieces of imaging data with medium-angle lighting, and multiple pieces of imaging data with high-angle lighting.

The judgment image generating section includes a maximum/minimum brightness image generating section and an excluded region specifying section.
The maximum/minimum luminance image generating unit synthesizes the plurality of low-angle illumination image data, the plurality of medium-angle illumination image data, and the plurality of high-angle illumination image data such that the maximum luminance value for each pixel position becomes the luminance value for that pixel position, thereby generating low-angle maximum luminance image data, medium-angle maximum luminance image data, and high-angle maximum luminance image data, and synthesizes the plurality of low-angle illumination image data, the plurality of medium-angle illumination image data, and the plurality of high-angle illumination image data such that the minimum luminance value for each pixel position becomes the luminance value for that pixel position, thereby generating low-angle minimum luminance image data, medium-angle minimum luminance image data, and high-angle minimum luminance image data.
The exclusion area identification unit identifies an exclusion target pixel area in an image represented by each of the low-angle minimum luminance image data, the medium-angle maximum luminance image data, the high-angle maximum luminance image data, the low-angle minimum luminance image data, the medium-angle minimum luminance image data, and the high-angle minimum luminance image data, based on at least one of the low-angle maximum luminance image data, the medium-angle minimum luminance image data, and the high-angle minimum luminance image data.

The pixel area to be excluded corresponds to an area included in the imaging area that is not to be inspected, and based on the low-angle minimum luminance image data, medium-angle minimum luminance image data, and high-angle minimum luminance image data that have been disabled for the pixel area to be excluded, low-angle judgment image data, medium-angle judgment image data, and high-angle judgment image data are generated as judgment image data, respectively.
The defect determination unit determines the presence or absence of defects in the imaging region other than the exclusion target pixel region based on the low angle determination image data, the medium angle determination image data, and the high angle determination image data.

The ceramic body defect inspection device is said to be able to reliably detect defects that should be detected without erroneously detecting irregularities on a normal ceramic surface as defects, and to perform defect inspection more efficiently by excluding areas that do not need to be inspected when generating image data for judgment.

WO2019/216031A1

However, cylindrical molded products such as gaskets and caps made of elastic materials can have molding defects such as uneven surfaces and cracks. Here, we propose an appearance inspection device that is more suitable for discovering molding defects in such molded products.

The appearance inspection device disclosed herein includes an inspection stage on which an object to be inspected is placed, at least one camera that photographs the object to be inspected placed on the inspection stage from the side of the object to be inspected, a first illuminator that irradiates the side surface of the object to be inspected placed on the inspection stage with light from above, a second illuminator that irradiates the side surface of the object to be inspected placed on the inspection stage with light from below, and a processing device. The processing device is configured to execute the following processes.
A process of acquiring a first image of an object to be inspected placed on the inspection stage by a camera with the second illumination turned off and the first illumination turned on. A process of acquiring a second image of an object to be inspected placed on the inspection stage by a camera with the first illumination turned off and the second illumination turned on. A judgment process of judging the quality of the photographed object to be inspected based on the first image and the second image.

The visual inspection device disclosed here can accurately extract defects that occur on the side surface of the object being inspected, allowing for stable pass/fail judgments.

FIG. 1 is a schematic side view of a visual inspection apparatus 10. As shown in FIG. FIG. 2 is a schematic plan view of the appearance inspection apparatus 10. As shown in FIG. FIG. 3 is a side view that shows a schematic diagram of a visual inspection apparatus 10A showing another embodiment. FIG. 4 is a photograph showing an example of the first image. FIG. 5 is a photograph showing an example of the second image. FIG. 6 shows an example of an image in which luminance values have been subjected to a synthesis operation.

Below, one embodiment of the invention disclosed herein will be described with reference to the drawings. Note that the present invention is not limited to the following embodiment. Each drawing is drawn diagrammatically and does not necessarily reflect the actual product. Furthermore, each drawing shows only an example, and does not limit the present invention unless otherwise specified. Furthermore, the same reference numerals are appropriately used for components and parts that perform the same function, and duplicate explanations will be omitted.

Figure 1 is a schematic side view of the appearance inspection device 10. Figure 2 is a schematic plan view of the appearance inspection device 10. As shown in Figures 1 and 2, the appearance inspection device 10 includes an inspection stage 11, a plurality of cameras 12, a first light 13, a second light 14, and a processing device 15.

<Test Subject 2>
In this embodiment, the test object 2 is a substantially cylindrical molded product. The test object 2 is made of an elastic material such as rubber or elastomer. The test object 2 has protrusions 2a and 2b protruding in the outer diameter direction at the upper and lower parts in the axial direction, which are continuously provided in the circumferential direction. In this way, the test object 2 may have protrusions protruding in the outer diameter direction formed at predetermined positions. In addition, several faces 2c are formed along the circumferential direction on the side peripheral surface, and ridge lines 2d are provided at the boundaries between the faces 2c. Many cylindrical molded products, such as gaskets inserted into syringes as medical rubber parts and caps attached to syringe nozzles, have such protrusions and ridge lines. The test object 2 shown in FIG. 1 and FIG. 2 is an example. The shape of the test object to be inspected by the appearance inspection device 10 is not limited to the shape shown in FIG. 1 and FIG. 2. In a shape having protrusions 2a, 2b and ridges 2d like this, when light is shone on it, the parts on which the light hits become bright due to the protrusions 2a, 2b and ridges 2d, and at the same time, shadows are cast behind them, making them dark.

<Inspection Stage 11>
The inspection stage 11 is a stage on which the object to be inspected 2 is placed. In this embodiment, since the object to be inspected 2 is a cylindrical molded product such as a gasket or a cap, the inspection stage 11 can be, for example, a circular pedestal on which the object to be inspected 2 is placed. In this embodiment, the inspection stage 11 uses a pedestal that is the same size as or slightly larger than the bottom surface of the object to be inspected 2. Furthermore, as long as the object to be inspected 2 can be placed stably, the upper surface 11a of the inspection stage 11 may be smaller than the bottom surface of the object to be inspected 2.

<Camera 12>
The camera 12 is arranged so as to photograph the object 2 placed on the inspection stage 11 from the side of the object 2. In this embodiment, as shown in FIG. 1 and FIG. 2, the appearance inspection device 10 includes a plurality of cameras 12. The plurality of cameras 12 are arranged so as to photograph the object 2 placed on the inspection stage 11 from a plurality of directions divided in the circumferential direction. The plurality of cameras 12 (six cameras in the example shown in FIG. 2) are arranged at predetermined intervals around the object 2 placed on the inspection stage 11. Here, for example, CA-H200M manufactured by Keyence Corporation may be used as the camera 12. Furthermore, ML-M3520UR manufactured by Moritex Corporation may be used as the lens associated with the camera 12. Furthermore, an intermediate ring may be used between the lens and the light receiving part of the camera. ANB84805 manufactured by Panasonic Corporation may be used as the intermediate ring. The intermediate ring may be used when the distance between the camera and the subject cannot be sufficiently secured due to the layout of the facility. The intermediate ring is also called the close-up ring.

<First lighting 13, second lighting 14>
The first illuminator 13 is an illuminator that irradiates light from above onto the position where the inspected object 2 is placed on the inspection stage 11. The second illuminator 14 is an illuminator that irradiates light from below onto the position where the inspected object 2 is placed on the inspection stage 11. The first illuminator 13 may be configured, for example, to irradiate illumination light or reflected light onto the side surface of the inspected object 2 from above. The second illuminator 14 may be configured, for example, to irradiate illumination light or reflected light onto the side surface of the inspected object 2 from below.

In the embodiment shown in FIG. 1, the first illumination 13 is composed of an upper ring illumination 41 (also referred to as the first ring illumination) that irradiates the position of the inspection object 2 on the inspection stage 11 from above with light. The second illumination 14 is composed of a lower ring illumination 42 (also referred to as the second ring illumination) that irradiates the position of the inspection object 2 on the inspection stage 11 from below with light. The upper ring illumination 41 and the lower ring illumination 42 are preferably configured to irradiate light at an approximately predetermined angle with respect to the position of the inspection object 2. By adopting the ring illuminations 41 and 42 for the first illumination 13 and the second illumination 14, respectively, uniform light is irradiated in the circumferential direction to the side surface of the inspection object 2 placed on the inspection stage 11. The first illumination 13 and the second illumination 14 are preferably connected to the processing device 15, and are preferably configured to control the timing of turning on and off the illumination.

Figure 3 is a schematic side view of a visual inspection device 10A showing another embodiment.

3, the second illumination 14 has a reflecting surface 51 and an illuminator 52. The reflecting surface 51 is provided around the position where the inspected object 2 is placed on the inspection stage 11, below the inspected object 2 placed on the inspection stage 11. The illuminator 52 can be a lighting device that reflects light on the reflecting surface 51 and irradiates the inspected object 2 from below. In this case, the reflecting surface 51 and the illuminator 52 are preferably arranged so that the light of the illuminator 52 is reflected and irradiated from below the inspected object 2. In this way, the reflecting surface 51 is preferably arranged below the inspected object 2 placed on the inspection stage 11, and the illuminator 52 may be arranged above the inspection stage 11. When the reflecting surface 51 is used in this way, even if another object is placed below the inspection stage 11 and an illuminator cannot be placed, the inspected object 2 can be configured to be irradiated with light from below.

<Reflecting surface 51>
In the embodiment shown in FIG. 3, the reflection surface 51 is provided at least around the position of the inspection stage 11 where the inspection object 2 is placed. In this embodiment, the inspection stage 11 is made of a white resin having a high reflectance overall, and the upper surface of the inspection stage 11 constitutes the reflection surface 51. In this case, the inspection stage 11 provided with the reflection surface 51 can be made of, for example, a resin base. Therefore, it can be constructed relatively inexpensively. A reflector made of a plate having a high reflectance may be attached to the upper surface of the inspection stage 11. In this way, the reflection surface 51 may be constructed on the upper surface of the inspection stage 11. The reflection surface 51 is not limited to such a form. The reflection surface 51 may be provided at an appropriate angle so that the light from the illuminator 52 is reflected toward the side surface of the inspection object 2. In addition, the reflection surface 51 may be disposed at an appropriate position so that the light from the illuminator 52 is reflected toward the side surface of the inspection object 2.

<Illuminator 52>
The illuminator 52 is arranged above the inspection stage 11 and is configured with a spot illuminator 61 that reflects light on the reflecting surface 51 and irradiates the inspection object 2 from below. In this case, a light-shielding masking 62 that blocks the center of the light of the spot illuminator 61 is further provided. In the embodiment shown in FIG. 3, the first ring illuminator 41 constituting the first illuminator 13 has a transparent plate 63 attached to the inside. The spot illuminator 61 is provided above the transparent plate 63. The light-shielding masking 62 is provided on the transparent plate 63. In this case, the light irradiated from the spot illuminator 61 is blocked at the center by the light-shielding masking 62. The light passes through the transparent plate 63 around the light-shielding masking 62 and is reflected by the reflecting surface 51 provided on the upper surface 11a of the inspection stage 11. The light reflected by the reflecting surface 51 illuminates the side peripheral surface of the inspection object 2 placed on the inspection stage 11 from below.

In this embodiment, a light-shielding mask 62 is placed directly above the object 2, and a spotlight 61 is provided directly above that. The light of the spotlight 61 is blocked at the center by the light-shielding mask 62, and passes through a transparent plate 63 around the light-shielding mask 62. A portion of the light that passes through the transparent plate 63 is reflected by the reflecting surface 51 and irradiated onto the side surface of the object 2. The side surface of the object 2 is irradiated with light that is approximately uniform in the circumferential direction. The position and shape of the mask 62 may be appropriately set depending on the position, size, shape, etc. of the object 2 to be inspected placed on the inspection stage 11. The transparent plate 63 may be a translucent plate as long as it transmits light.

In the embodiment shown in FIG. 3, the reflective surface 51 may be subjected to an appropriate surface treatment to suppress the diffusion of light. In the embodiment shown in FIG. 3, the spot light 61 installed above is an LDR2-50SW2 manufactured by CCS Inc. In addition, the ring light 41 installed below the spot light 61 may be an HPR2-150SW manufactured by CCS Inc.

Although not shown, the appearance inspection device is not limited to the above-mentioned embodiment. The configuration of the first illumination 13 that irradiates the side surface of the inspected object 2 placed on the inspection stage 11 with light from above, and the second illumination 14 that irradiates the side surface of the inspected object 2 placed on the inspection stage 11 with light from below is not limited to the above-mentioned embodiment, and can be changed to various forms. For example, the illuminator 52 of the second illumination 14 may be a panel illumination. In this case, the panel illumination may be arranged so that light is irradiated downward inside (inner diameter side) of the first ring illumination 41 that constitutes the first illumination 13. A light-shielding masking may be attached to the center of the panel illumination as the second illumination 14. Alternatively, the panel illumination as the second illumination 14 may be a ring-shaped panel illumination. If the panel illumination is a ring-shaped panel illumination, the light-shielding masking may not be required. In this configuration, the light from the panel illumination as the second illumination 14 is reflected by the reflecting surface 51 of the inspection stage 11 and irradiated onto the side surface of the inspection object 2 placed on the inspection stage 11. In this case, the light reflected from the panel illumination by the reflecting surface 51 is irradiated onto the side surface of the inspection object 2, and the side surface of the inspection object 2 is irradiated with light that is approximately uniform in the circumferential direction.

In this way, various examples have been given of the configurations of the first illumination 13 and the second illumination 14 of the appearance inspection device 10. The first illumination 13 may be configured to irradiate the side of the object to be inspected from above with illumination light or reflected light. The second illumination 14 may be configured to irradiate the side of the object to be inspected from below with illumination light or reflected light. In this respect, the present invention is not limited to the above-mentioned embodiment. Here, the first illumination 13 and the second illumination 14 may have different wavelengths of light. For example, when inspecting minute irregularities, it is preferable to use illumination with a short wavelength (blue). When illumination with a short wavelength (blue) is used, it is expected that the resolution will be high. On the other hand, when the subject base material itself has fine irregularities, illumination with a long wavelength (red) may be used to reduce the influence of the unevenness of the base material. In this way, illumination with different wavelengths of light may be used according to the characteristics of the object to be inspected 2. In addition, a polarized lens may be attached to at least one of the light-emitting surfaces of the first illumination 13 and the second illumination 14 and the lens tip of the camera 12. In this case, the reflected light of the illumination reflected on the inspection subject 2 is removed using a polarizing lens. Attaching a polarizing lens can be effective for inspecting an inspection subject 2 whose surface is covered with a film or the like and which is prone to reflecting light.

Processing device 15
The processing device 15 is a device that performs various processes of the appearance inspection device 10. The processing device 15 can be embodied by a computer that runs according to a predetermined program. Specifically, each function of the processing device 15 is processed by an arithmetic device (also called a processor, CPU (Central Processing Unit), or MPU (Micro-processing unit)) or a storage device (such as a memory or a hard disk) of each computer that constitutes the processing device 15. For example, each configuration of the processing device 15 can be embodied as a database that stores data embodied by a computer in a predetermined format, a data structure, a processing module that performs a predetermined arithmetic process according to a predetermined program, or as a part of them. Although not shown in the figure, the processing device 15 may be one in which multiple control devices work together. For example, when the processing device 15 of the appearance inspection device 10 is connected to an external computer so as to be able to perform data communication through a LAN cable, the Internet, or the like, the processing of the processing device 15 may be performed in cooperation with such an external computer. For example, the information or a part of the information stored in the processing device 15 may be stored by an external computer, and the processing or a part of the processing performed by the processing device 15 may be performed by an external computer.

The processing device 15 is configured to perform a process 31 for acquiring a first image, a process 32 for acquiring a second image, and a judgment process 33 for judging pass/fail. In this embodiment, the processing device 15 is configured to operate the camera 12, the first lighting 13, and the second lighting 14 according to a predetermined program. The processing device 15 is electrically connected to the camera 12, the first lighting 13, and the second lighting 14 through a communication network. The communication network may be either a wired or wireless network. The first and second images are captured by the processing device 15 appropriately operating the first lighting 13, the second lighting 14, and the camera 12.

<First image>
The first image is an image of the inspection object 2 placed on the inspection stage 11 photographed by the camera 12 with the second illumination 14 turned off and the first illumination 13 turned on. In the process 31 of acquiring the first image, it is preferable that the inspection object 2 placed on the inspection stage 11 is photographed by the camera 12 with the second illumination 14 turned off and the first illumination 13 turned on.

Figure 4 is a photograph showing an example of the first image. The test object 2 shown in the photograph of Figure 4 has a shape roughly similar to that shown in Figure 1. The test object 2 is a molded product having a roughly cylindrical shape. The test object 2 has protrusions 2a and 2b that protrude in the outer diameter direction at the upper and lower parts in the axial direction, which are continuously provided in the circumferential direction. Several faces 2c are formed along the circumferential direction on the side peripheral surface, and a ridge line 2d is provided at the boundary between the faces 2c. As shown in Figure 4, when the second illumination 14 is turned off and the first illumination 13 is irradiated, and the test object 2 placed on the inspection stage 11 is photographed by the camera 12, the light of the first illumination 13 is likely to hit the upper side of the protrusion on the side peripheral surface of the test object 2. In addition, a shadow due to the light of the first illumination 13 is likely to be generated on the lower side of the protrusion. In the example of the first image shown in Figure 4, the upper side of the protrusion 2b at the lower part of the test object 2 is likely to be hit by the light of the first illumination 13, and has a high brightness. In addition, the area below the protrusion 2a at the top of the test subject 2 is prone to shadows and has low brightness.

<Second image>
The second image is an image of the inspection object 2 placed on the inspection stage 11 photographed by the camera 12 with the first illumination 13 turned off and the second illumination 14 turned on. In the process 32 of acquiring the second image, it is preferable that the inspection object 2 placed on the inspection stage 11 is photographed by the camera 12 with the first illumination 13 turned off and the second illumination 14 turned on.

Figure 5 is a photograph showing an example of the second image. The test object 2 shown in the photograph of Figure 5 is the same as the test object 2 in Figure 4. As shown in Figure 5, when the test object 2 placed on the inspection stage 11 is photographed by the camera 12 with the first illumination 13 turned off and the second illumination 14 illuminated, the light of the second illumination 14 is likely to fall on the underside of the protrusion on the side circumference of the test object 2. Also, a shadow caused by the light of the second illumination 14 is likely to be cast on the upper side of the protrusion. For this reason, in the example of the second image shown in Figure 5, the underside of the protrusion 2b at the top of the test object 2 is likely to be struck by the light of the second illumination 14, and has a high brightness. Also, the underside of the protrusion 2a at the top of the test object 2 is likely to be cast by a shadow, and has a low brightness.

<Determination Process>
The judgment process is a process for judging the quality of the photographed inspection object based on the first image and the second image.

For example, if there is a dent on the side surface of the test object 2, in the first image, a shadow will be cast on the top of the dent, resulting in a low brightness, and the brightness of the bottom of the dent will tend to be high. Also, in the second image, a shadow will be cast on the bottom of the dent, resulting in a low brightness, and the brightness of the top of the dent will tend to be high. Also, if there is a raised portion on the side surface of the test object 2, in the first image, the brightness of the top of the raised portion will be high, and the brightness of the bottom of the raised portion will be low. Also, in the second image, the brightness of the bottom of the raised portion will be high, and the brightness of the bottom of the raised portion will be low. A dent or raised portion at a predetermined position on a molded product is normal, but if there is a dent or raised portion at a position other than the predetermined position, there is a possibility of molding defects. In this way, the first and second images are acquired, and by determining whether there is a dent or raised portion at a position other than the predetermined position, it is possible to determine whether the photographed test object is good or bad.

The side surface of the test object 2 shown in FIG. 1 has several faces 2c formed along the circumferential direction, and ridges 2d are provided at the boundaries between the faces 2c. In this embodiment, as shown in FIG. 2, a plurality of cameras 12 are arranged at predetermined intervals around the inspection stage 11 so that the test object 2 is photographed from a plurality of directions divided in the circumferential direction of the inspection stage 11. The plurality of cameras 12 are arranged so that, as a whole, the side surface of the test object 2 placed on the inspection stage 11 can be photographed over the entire circumference. Therefore, the faces 2c and ridges 2d formed on the side surface of the test object 2 are also captured in one of the images. In this case, if the faces 2c and ridges 2d have recesses or raised portions, the faces 2c will have portions with changed brightness, and the ridges 2d will have portions with distorted brightness.

1 and 2, the appearance inspection device 10 includes a plurality of cameras 12 that take images of the side of the object 2 placed on the inspection stage 11, a first illuminator 13 that irradiates the side surface of the object 2 from above, and a second illuminator 14 that irradiates the side surface of the object 2 from below. The processing device 15 is configured to determine the quality of the object 2 based on a first image taken of the object 2 with the second illuminator 14 turned off and the first illuminator 13 turned on, and a second image taken of the object 2 with the first illuminator 13 turned off and the second illuminator 14 turned on. In this case, the first illuminator 13 and the second illuminator 14 apply light to the side surface of the object 2 from above and below, respectively. Then, a first image is obtained by shining light from above onto the side surface of the test object 2 using the first illumination 13, and a second image is obtained by shining light from below.

As a result, if there is a concave-convex molding defect or a scratch on the side surface of the inspected object 2, it is possible to create vertical shadows. Furthermore, shadows caused by the product's inherent undulations can also be inspected by using the image of the side where light hits the undulations. In this way, it is possible to determine whether the photographed side surface of the inspected object 2 is good or bad based on the first image obtained by shining light from above on the side surface of the inspected object 2 using the first illumination 13, and the second image obtained by shining light from below.

In addition, in this embodiment, as described above, the first illuminator 13 and the second illuminator 14 each illuminate the side surface of the inspection object 2 with light of uniform illuminance in the circumferential direction. Therefore, in the first image and the second image, the brightness of the areas illuminated by the light and the shadow of the shaded areas are approximately the same. Therefore, it is possible to accurately determine whether the side surface of the photographed inspection object 2 is good or bad.

In addition, in this embodiment, multiple cameras 12 are arranged so that the object 2 placed on the inspection stage 11 is photographed from multiple directions divided in the circumferential direction. Therefore, the entire circumference of the side surface of the object 2 placed on the inspection stage 11 can be photographed without rotating the object 2 placed on the inspection stage 11. Since the entire circumference of the side surface of the object 2 placed on the inspection stage 11 can be photographed without rotating the object 2 placed on the inspection stage 11, there is no misalignment of the rotation. This allows for stable inspection and allows for accurate identification of defective areas. Furthermore, since multiple cameras 12 are used, the time required to photograph the entire circumference of the side surface of the object 2 is short.

In addition, the first light 13 and the second light 14 irradiate the side surface of the test object 2 from above and below. The camera 12 photographs the side surface of the test object 2 from the radial direction. Therefore, even if the test object 2 has a cylindrical shape, the variation in brightness value in the circumferential direction is suppressed to be small. The inspection area per camera 12 is wider than when light is applied from the horizontal direction of the test object 2. Therefore, even if multiple cameras 12 are arranged along the circumferential direction of the inspection stage 11, the number of cameras 12 required is small, and the equipment cost is suppressed to be small. For example, in this embodiment, six cameras 12 cover an area divided into six circumferential areas of the test object 2. The number of cameras 12 is not limited to six, and may be even less, such as five, four, or three.

In the judgment process, the quality of the photographed test object may be judged based on a process of removing the shading gradient of the photographed test object in the first image, a process of removing the shading gradient of the photographed test object in the second image, and the first and second images from which the shading gradient has been removed. For example, as shown in FIG. 1, when the test object 2 is a member having an approximately cylindrical shape, even if light is evenly applied in the circumferential direction, the part closer to the illumination tends to appear brighter, and the part farther from the illumination tends to appear darker. In addition, when the illuminance of the light irradiated to the test object 2 is uneven in the circumferential direction, uneven brightness may occur in the circumferential direction of the test object 2. In such a case, a certain shading gradient tends to occur in the photographed test object 2 in the circumferential direction of the test object 2. For this reason, the shading can be removed by performing a process of removing the shading gradient of the photographed test object in the first image and a process of removing the shading gradient of the photographed test object in the second image. For example, FIG. 4 and FIG. 5 are photographs before the shading is removed. In this case, as shown in Figures 4 and 5, the grayscale gradient tends to be stronger in the height direction than in the circumferential direction. The process of removing the grayscale gradient can also remove this tendency. By performing the process of removing the grayscale gradient of the test object 2 in the first and second images, the grayscale gradient in the height and circumferential directions of the test object 2 is removed, which can improve the accuracy of determining whether the photographed test object 2 is good or bad.

The process of removing the shading gradient here can be realized by, for example, a process of removing shading with a real-time shading correction filter in commercially available image processing software such as XG8500 manufactured by Keyence Corporation. With this process, the gradual change in shading (shading) on the surface of the photographed test object 2 is removed by calculating the difference between the input image and an internally processed image created from the input image. By removing the change in shading on the surface of the photographed test object 2, the parts with abrupt changes in contrast are emphasized and extracted. In addition, the contrast of the side surface of the photographed test object 2 may be uniformly corrected using the average or median density within the region.

The judgment process may be configured, for example, to calculate the difference in luminance value between adjacent pixels or groups of pixels in a first image from which the grayscale gradient has been removed and a second image from which the grayscale gradient has been removed, and if the difference is equal to or greater than a threshold, it is judged to be a defective part. When there are no molding defects such as scratches, the difference in luminance value between adjacent pixels or groups of pixels vertically or horizontally is roughly constant. In a first image from which the grayscale gradient has been removed and a second image from which the grayscale gradient has been removed, the difference in luminance value between adjacent pixels or groups of pixels vertically or horizontally is calculated, and if the difference is equal to or greater than a threshold, it is suspected that the part is defective. For this reason, it may be configured to be judged to be a defective part.

Note that here, the pixel group means a group of pixels in an arbitrarily determined area that is regarded as one block. In other words, instead of evaluating the difference for each pixel, a plurality of pixels in an area where a plurality of pixels are gathered may be treated as one pixel group, and the difference may be evaluated for the pixel group. This allows the surface condition of the inspected object 2 to be evaluated in an area of appropriate size that is not too small and not too subdivided. The size of the area that defines the pixel group should be set to an appropriate size for evaluating the surface condition of the inspected object 2. For example, a rectangular pixel area such as 4 (2 x 2), 9 (3 x 3), or 16 (4 x 4) should be set. By calculating the difference in brightness value for each pixel group of appropriate size in this way, the possibility of false detection of molding defects, scratches, etc. is reduced. In addition, the brightness value of the pixel group may be evaluated, for example, as the average brightness value of the pixels included in the pixel group.

The processing device 15 may include a process for obtaining an image obtained by performing a composite operation on the luminance values of the first image from which the grayscale gradient has been removed and the second image from which the grayscale gradient has been removed. In this case, the judgment process may include a process for judging the quality of the photographed test object 2 based on the image from which the luminance values have been compositely calculated. When there are no molding defects such as scratches, the difference in luminance values between adjacent pixels or pixel groups vertically or horizontally is approximately constant. In a normal product, the image obtained by performing a composite operation on the luminance values of the first image from which the grayscale gradient has been removed and the second image from which the grayscale gradient has been removed will have an approximately constant pattern. In other words, the normal inherent undulations, ridges, protrusions, etc. of a molded product are obtained in a constant pattern of luminance value differences. For this reason, by obtaining an image obtained by performing a composite operation on the luminance values of the first image from which the grayscale gradient has been removed and the second image from which the grayscale gradient has been removed, and taking the difference from a pattern of luminance values of a normal product prepared in advance, a portion with a certain or greater difference can be extracted as a suspected defective portion. When a suspected defective portion is extracted, it may be configured to be judged as a defective portion.

Here, the "composite operation" may employ an appropriate calculation method for evaluating the surface condition of the inspection object 2 and extracting defective parts using the first image from which the grayscale gradient has been removed and the second image from which the grayscale gradient has been removed. In other words, performing a composite operation on the luminance values includes various processes for combining the luminance values using a predetermined arithmetic formula. Here, the composite operation may involve, for example, simply adding up the luminance values. Furthermore, the composite operation is not limited to simply adding up the luminance values, and may involve subtraction, averaging, addition or subtraction of absolute values, etc. The method of the composite operation may be determined by performing tests in advance and determining an appropriate calculation method for extracting defective parts.

For example, FIG. 6 shows an image in which the brightness values have been combined. FIG. 6 shows an image obtained by the difference in brightness values between the first image shown in FIG. 4 and the second image shown in FIG. 5. In the first image shown in FIG. 4 and the second image shown in FIG. 5, there is a defective part with a gentle bulge in the blue enclosure A. Such an abnormality may appear normal depending on the material of the test object 2 and the way the light hits it, so it may be difficult to distinguish even with the naked eye. Even in this case, as described above, the first image illuminated from above and the second image illuminated from below are obtained, and the grayscale gradient is removed from each. Furthermore, according to the image in which the difference in brightness values has been combined, the parts with different contrasts are emphasized as shown in FIG. 6. This makes it easier to extract defective parts in computer image processing.

In the determination process, if the difference in luminance value between adjacent pixels or groups of pixels in an image obtained by combining the luminance values of the first image and the second image is equal to or greater than a predetermined threshold value, the pixel or group of pixels may be determined to be a defective part. For example, if there is a scratch on the side surface of the test object 2, the difference in luminance value between adjacent pixels or groups of pixels will be clearly visible. For this reason, it becomes easy to extract scratches on the side surface of the test object 2 by determining whether the difference in luminance value between adjacent pixels or groups of pixels vertically or horizontally is equal to or greater than a predetermined threshold value.

The processing device 15 may store in advance reference data indicating the distribution of luminance values in multiple areas for an image obtained by combining the luminance values of the first image and the second image when the inspected object 2 is normal. In this case, the judgment process may be configured to extract defective areas based on the distribution of luminance values in the image obtained by combining the luminance values of the first image and the second image and the reference data.

The processing device 15 may also include a display 40 as shown in FIG. 1. The processing device 15 may be configured to display the color of each pixel in the multiple areas in a predetermined color according to the difference in brightness between the distribution of brightness values in an image obtained by synthesizing the brightness values of the first image and the second image and the reference data. In this case, the defective parts are clearly indicated to the user through the image displayed on the display 40.

The processing device 15 may further be configured to use a machine learning model such as a neural network to determine the quality of the photographed specimen 2. In this case, it is preferable that the first and second images of the specimen 2 and data indicating the positions of defects such as scratches on the photographed specimen 2 that appear in the first and second images are prepared as learning data. It is preferable that a trained machine learning model is prepared so that a quality determination of the photographed specimen 2 is output as a result based on the first and second images of the specimen 2 by learning such learning data. A neural network may be used as the machine learning model.

If the result of the pass/fail judgment is a defect, the defective part of the photographed test object 2 may be further identified. The defective part of the test object 2 may be identified, for example, by outputting position information, color-coding the defective part on the display, or by being surrounded by a figure such as a circle. This makes it easier for the user to understand the position of the defective part.

In this case, for example, the processing device 15 may be equipped with a machine learning model configured to learn the first image, the second image, and the position of the flaw in the photographed object as learning data, and to determine whether the photographed object is good or bad based on the first image and the second image.

The processing device 15 may also be equipped with a machine learning model configured to learn the first and second images from which the shading gradient has been removed and the location of scratches on the photographed test object as learning data, and to determine the quality of the photographed test object based on the first and second images from which the shading gradient has been removed.

The processing device 15 may be equipped with a machine learning model configured to learn the image obtained by performing a composite operation on the brightness values and the position of the flaw in the photographed object to be inspected as learning data, and to judge the quality of the photographed object to be inspected based on the image obtained by performing a composite operation on the brightness values. Examples of the machine learning model are not limited to those exemplified here.

As described above, the appearance inspection device 10 includes the inspection stage 11, the camera 12, the first lighting 13, the second lighting 14, and the processing device 15. The inspection stage 11 is a base on which the inspected object 2 is placed. The camera 12 is an imaging device that images the inspected object 2 placed on the inspection stage 11 from the side of the inspected object 2. The first lighting 13 is an illumination that irradiates the side surface of the inspected object 2 placed on the inspection stage 11 with light from above. The first lighting 13 is an illumination that irradiates the side surface of the inspected object 2 placed on the inspection stage with light from below.

In the appearance inspection apparatus 10, the following processes 31 to 33 are executed (see FIG. 1).
Process 31: A process of acquiring a first image of the inspection object 2 placed on the inspection stage 11 by the camera 12 with the second illumination 14 turned off and the first illumination 13 turned on. Process 32: A process of acquiring a second image of the inspection object 2 placed on the inspection stage 11 by the camera 12 with the first illumination 13 turned off and the second illumination 14 turned on. Process 33: A judgment process of judging the quality of the photographed inspection object 2 based on the first image and the second image.

The appearance inspection device 10 can accurately extract defects that occur on the side surface of the inspected object 2, such as uneven defects such as dents and bulges, cracks, tapered distortion of protrusions, warping, and other defects that are specific to normal products, allowing for stable pass/fail judgments.

In the process of obtaining the first and second images, the above-described embodiment includes a plurality of cameras 12 that capture images of the inspected object 2 placed on the inspection stage 11 from a plurality of directions divided in the circumferential direction (see FIG. 1). Unless otherwise specified, in the appearance inspection device 10, the camera 12 may be movable relative to the inspection stage 11. In this case, the camera 12 may be configured as a single unit. The inspection stage 11 may rotate with respect to the inspection stage 11. The camera 12 may also be configured to rotate around the inspection stage 11.

Although the invention disclosed herein has been described in detail above, the invention disclosed herein is not limited to the above-mentioned embodiments and modifications unless otherwise specified. Furthermore, the configurations of the various embodiments and modifications mentioned above can be combined as appropriate as long as they do not interfere with each other.

Reference Signs List 2 Object to be inspected 2a, 2b Protrusion 2d Ridge line 10, 10A Appearance inspection device 11 Inspection stage 12 Camera 13 First illumination 14 Second illumination 15 Processing device 31 Process for acquiring first image 32 Process for acquiring second image 33 Judgment process 41 Upper ring illumination (first ring illumination)
42 Lower ring illumination (second ring illumination)
51 Reflective surface 52 Illuminator 61 Spot lighting 62 Masking 63 Transparent plate

Claims (7)

an inspection stage on which an object to be inspected is placed and on which a reflecting surface is provided at least around the periphery of the position on which the object to be inspected is placed ;
At least one camera that captures an image of the object to be inspected placed on the inspection stage from a side of the object to be inspected;
a first illuminator which is a ring illuminator that irradiates a side peripheral surface of the inspection object placed on the inspection stage with light from above;
a second illumination device that includes a spot illumination device disposed above the inspection stage and that illuminates a side peripheral surface of the inspection object placed on the inspection stage from below by reflecting light on the reflecting surface ;
a processing device;
The processing device includes:
a process of acquiring a first image of an object to be inspected placed on the inspection stage by the camera in a state in which the second illumination is turned off and the first illumination is applied;
a process of acquiring a second image of an object to be inspected placed on the inspection stage by the camera in a state in which the first illumination is turned off and the second illumination is applied;
and performing a judgment process for judging the quality of the photographed inspection object based on the first image and the second image. 2. The visual inspection apparatus according to claim 1, further comprising a light-shielding masking disposed directly above the object to be inspected placed on said inspection stage. 3. An appearance inspection apparatus as described in claim 2, wherein a light-transmitting plate is placed directly above the object to be inspected placed on the inspection stage, the light-shielding masking is provided on the light-transmitting plate, and the second lighting consisting of the spot lighting is placed directly above the masking. 4. The visual inspection device according to claim 1, wherein the reflecting surface is subjected to a surface treatment for suppressing diffusion of light. The determination process includes:
A process of removing a gray gradient of the photographed object from the first image;
A process of removing a gray gradient of the photographed object from the second image;
An appearance inspection device as described in any one of claims 1 to 4, which performs a judgment process to judge the quality of the photographed inspection object based on the first image and the second image from which the gradation gradient has been removed. The determination process calculates a difference in luminance between adjacent pixels or groups of pixels in a first image from which a gradient has been removed and a second image from which a gradient has been removed;
If the difference is equal to or greater than the threshold, it is determined to be a defective part.
6. The visual inspection apparatus according to claim 5 . The processing device includes:
The method includes a process of obtaining an image by performing a synthesis operation on the luminance values of the first image and the second image from which the gray scale gradient has been removed, and
The determination process includes:
In an image obtained by synthesizing the luminance values of the first image and the second image, if a difference in luminance value between adjacent pixels or pixel groups is equal to or greater than a predetermined threshold value, the pixel or pixel group is determined to be a defective portion.
7. An appearance inspection apparatus according to claim 5 or 6 .

Images