Inspection device and inspection method
The inspection apparatus and method address inefficiencies in defect detection on glossy curved surfaces by using a polarization-based system to generate spatial frequency optimized images, improving accuracy and reducing costs by eliminating the need for multiple optical systems and inspectors.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- SONY GROUP CORP
- Filing Date
- 2022-01-06
- Publication Date
- 2026-07-22
Smart Images
Figure 0007893248000004 
Figure 0007893248000005 
Figure 0007893248000006
Abstract
Description
Technical Field
[0001] The present technology relates to an inspection apparatus and an inspection method for inspecting the surface of an inspection target surface.
Background Art
[0002] Conventionally, for example, the defect inspection of the painted surface of a glossy curved surface such as an automobile body has been performed by visual inspection by an inspector. At this time, the inspector changes his / her observation position with respect to the illumination to change the illumination conditions, and detects defects such as scratches and unevenness.
[0003] In such defect inspection by visual inspection by an inspector, there may be variations in inspection accuracy depending on the inspector. In addition, for the inspection of large automobiles, a plurality of inspectors are required, which may increase the labor cost and become a factor in raising the production cost. Therefore, automation of defect inspection has been desired, and in recent years, the development of inspection apparatuses for optically inspecting surface defects has been promoted.
[0004] Patent Document 1 describes optically detecting surface defects of organic thin films such as polarizing films used in organic thin film solar cells, organic EL (Electro Luminescence) displays, liquid crystal displays, and the like.
[0005] In addition, when inspecting a plurality of types of defects, suitable illumination conditions differ depending on the characteristics of the defects. For example, in the detection of scratches, a method of dark field observation for detecting scattered light at the defective portion is effective. On the other hand, in the detection of dents, a method of using linear illumination such as a fluorescent lamp to visualize the change in the shape of the surface is effective. Therefore, a plurality of optical systems are required, and in order to realize a wide range of inspections, a large number of cameras and illuminations need to be installed.
Prior Art Documents
Patent Documents
[0006]
Patent Document 1
[0007] In detecting defects on a surface under inspection, there is a need for a method that can detect multiple types of defects over a wide area.
[0008] In light of the above circumstances, the objective of this technology is to provide an inspection device and inspection method suitable for inspecting the surface condition of a surface to be inspected. [Means for solving the problem]
[0009] The inspection apparatus relating to this technology comprises an irradiation unit, a polarization separation unit, an imaging unit, and a processing unit. The above-mentioned irradiation unit irradiates light onto the surface to be inspected. The polarization separation unit separates the light obtained from the surface to be inspected, which has been irradiated with the light, into multiple polarization components with different polarization directions. The imaging unit has multiple pixels that receive light with multiple different polarization components separated by the polarization separation unit and output pixel signals. The above processing unit filters the polarization phase difference image generated using the pixel signal output from the imaging unit at a predetermined spatial frequency.
[0010] This configuration makes it possible to detect multiple different types of defects on the surface being inspected.
[0011] The surface to be inspected may have a curved surface. The area of the light-emitting surface of the irradiation unit may be set to be larger than the surface to be inspected. The above processing unit may generate a polarization degree image using the pixel signals output from the above imaging unit.
[0012] The polarization separation unit has a plurality of polarizers that separate the light obtained from the surface to be inspected into polarization components with different polarization directions, and each of the plurality of polarizers may be arranged on the light-receiving surface of the corresponding pixel. Each of the above polarizers may have polarization axes at angles of α degrees, (α + 45) degrees, (α + 90) degrees, and (α + 135) degrees.
[0013] The above imaging unit may have a shine-proof optical system.
[0014] The inspection method according to the present technology is obtaining the pixel signal from an imaging unit having a plurality of pixels that receives light obtained from an inspection target surface irradiated with light through a polarization separation unit that separates the light into a plurality of different polarization components and outputs a pixel signal, and filtering the polarization phase difference image generated using the pixel signal at a predetermined spatial frequency.
Brief Description of the Drawings
[0015] [Figure 1] It is a schematic diagram showing the overall configuration of an inspection apparatus according to an embodiment of the present technology. [Figure 2] It is a schematic configuration diagram of an irradiation unit that constitutes a part of the above inspection apparatus. [Figure 3] It is a schematic configuration diagram of a polarization camera that constitutes a part of the above inspection apparatus. [Figure 4] It is a schematic configuration diagram of a polarization camera as another example. [Figure 5] It is an example of an image acquired by the above inspection apparatus. [Figure 6] It is a flowchart of an inspection method by the processing unit of the above inspection apparatus.
Modes for Carrying Out the Invention
[0016] Hereinafter, embodiments according to the present technology will be described with reference to the drawings. The inspection apparatus of this embodiment can be used, for example, for defect inspection of a glossy painted surface of a vehicle body (hereinafter, may be simply referred to as "painted surface"). The painted surface constituting the outer surface of the vehicle body is mainly composed of resin. The inspection apparatus is provided, for example, on a vehicle body painting line in an automobile manufacturing factory.
[0017] Defects on the painted surface of the vehicle body include, for example, convex defects, concave defects, etc. Convex defects are, for example, bumps and sags. A bump is a phenomenon in which the coating film bulges (convex) due to foreign substances such as fibers and sand dust mixed in the paint or foreign substances attached after painting. Sag refers to the situation where the paint flows downward before drying during painting on a vertical or inclined surface, resulting in a partially uneven coating film thickness. The light reflected at the positions of bumps and sags tends to have fewer scattered reflection components. Concave defects are, for example, peeling, dents, scratches, etc. Peeling is a hole where the paint peels off and reaches the base surface of the object to be painted from the coating film surface. A dent is a depression generated on the coating film surface where the paint peels off. A scratch is a linear scratch like a scuff mark. The light reflected at the position of the scratch has a scattered reflection component as the main component, while the light reflected at the positions of peeling and dents tends to have fewer scattered reflection components.
[0018] [Configuration of the inspection device] FIG. 1 is a schematic diagram showing the overall configuration of the inspection device 1. The inspection device 1 inspects defects present on the painted surface of the vehicle body M, which is the inspection target surface 2. In the example shown in FIG. 1, the inspection target surface 2 is the painted surface of the bonnet of the vehicle body M. The inspection device 1 includes an irradiation unit 5, a polarization camera 3, a processing unit 4, and a display unit 6. In the defect inspection by the inspection device 1, the inspection target surface 2 is imaged by the polarization camera 3 while the inspection target surface 2 is irradiated with light from the irradiation unit 5. In the imaging result, the defect inspection of the painted surface is performed by observing the change in the state of light (change in polarization state) due to reflection on the inspection target surface 2. This will be described in detail below.
[0019] (Irradiation unit) The irradiation unit 5 is surface illumination that irradiates the inspection target surface 2 with light. The light emitted from the irradiation unit 5 may be polarized light with a known polarization state such as circularly polarized light or linearly polarized light, or may be unpolarized light. Note that elliptically polarized light is also included in circularly polarized light. The light from the irradiation unit 5 is incident obliquely on the inspection target surface 2.
[0020] Figures 2(A) to 2(C) each show examples of the configuration of the irradiation unit 5. As shown in Figure 2(A), when the irradiation unit 5 emits unpolarized light, the irradiation unit 5 has, for example, a light source 50. The light source 50 is, for example, a white light source. As shown in Figure 2(B), when the irradiation unit 5 emits linearly polarized light, the irradiation unit 5 comprises a light source 50 and a polarizing plate 51. The polarizing plate 51 converts the unpolarized light from the light source 50 into linearly polarized light. As shown in Figure 2(C), when the irradiation unit 5 emits circularly polarized light, the irradiation unit 5 includes a light source 50, a polarizing plate 51, and a quarter-wave plate 52. The polarizing plate 51 converts the unpolarized light emitted from the light source 50 into linearly polarized light, and the quarter-wave plate 52 converts the linearly polarized light into circularly polarized light. The light source 50 described above is configured in a panel shape, for example, by placing a diffuser plate on an LED substrate in which multiple LEDs are arranged at equal intervals on a wiring board. Alternatively, the light source 50 may be configured by placing a diffuser plate on a group of fluorescent lamps arranged in a row, and is not particularly limited.
[0021] Here, the body of the vehicle M, which is the surface to be inspected, is typically not composed solely of flat surfaces with zero curvature, but rather has surfaces with curvature (curved surfaces). Furthermore, the curvature of the vehicle M body varies from part to part, and the inclination of the surfaces also varies from part to part. Here, "having curvature" means that the curvature is other than zero.
[0022] Considering the direction of light reflection due to the curved shape of the vehicle body M, it is preferable to set the area of the light-emitting surface of the illumination unit 5 to be sufficiently large. By configuring the area of the light-emitting surface of the illumination unit 5 to be large so that the polarization camera 3 is positioned on the optical axis of the light emitted from the illumination unit 5 and specularly reflected (reflected) by the inspection target surface 2, it becomes possible to inspect the curved surface with a specular reflection optical system, even if the inspection target surface 2 has a curved surface. In this way, by making the light-emitting surface of the illumination unit 5 large, the illumination unit 5 and the polarization camera 3 can be arranged in a specular reflection configuration.
[0023] This technology detects defects on the inspection surface 2 by observing the change in polarization state due to light reflection on the inspection surface 2. Therefore, by making the light-emitting surface of the irradiation unit 5 large and arranging the irradiation unit 5 and the polarization camera 3 in a specular reflection configuration, it becomes possible to inspect the curved surface of the inspection surface 2 with a specular reflection optical system, even if the inspection surface 2 has a curved surface, thereby increasing the accuracy of defect detection. In the example shown in Figure 1, where the hood is the inspection surface, if the width dimension of the vehicle body M is 170 cm, the horizontal and vertical dimensions of the light-emitting surface of the illumination unit 5 can be approximately 200 cm to 250 cm, respectively. In this way, the size of the light-emitting surface of the illumination unit 5 can be made larger than the inspection surface. Note that the numerical values are just examples and are not limited to them.
[0024] Furthermore, since this technology uses polarization for inspection, the highest sensitivity is obtained when the illumination unit and the polarization camera (imaging unit) are positioned at an angle centered on the Brewster angle. However, depending on the equipment arrangement, it is possible to achieve the desired result even at angles other than the Brewster angle.
[0025] (Polarization camera) Polarization camera 3 can acquire polarization information from the subject (surface to be inspected). The number of polarization cameras 3 installed is set appropriately according to the size of the surface to be inspected, and one or more cameras are installed. Figures 3 and 4 are schematic diagrams showing example structures of the polarization camera 3, respectively. Either the polarization camera configuration shown in Figure 3 or Figure 4 may be used.
[0026] Figure 3(A) is a schematic diagram showing an example of a polarizing camera 3. As shown in Figure 3(A), the polarization camera 3 has an image sensor 32 as an imaging unit and a polarization unit 31 as a polarization separation unit. The polarization unit 31 has a plurality of polarizers 34. Here, a configuration without a color filter is described, but the polarization camera 3 may have a color filter.
[0027] Figure 3(B) is a schematic diagram showing an example of an array of light-receiving modifiers in which polarizers 34 are arranged corresponding to pixels 33 of an image sensor 32. In Figure 3(B), multiple pixels 33 on which polarizers 34 (polarization units 31) are arranged are schematically illustrated.
[0028] The image sensor 32 has multiple pixels 33, each capable of outputting a pixel signal. As shown in Figure 3(B), the multiple pixels 33 are arranged in a two-dimensional matrix along two mutually orthogonal directions (X direction and Y direction) on the light-receiving surface of the image sensor 32. The image sensor 32 detects the intensity of incident light incident on each pixel 33, and the detection result is output as a pixel signal.
[0029] The specific configuration of the image sensor 32 is not limited; for example, a CMOS (Complementary Metal-Oxide Semiconductor) sensor or a CCD (Charge Coupled Device) sensor may be used as appropriate.
[0030] The polarization unit 31 separates the light (reflected light) obtained from the inspection target surface 2, which is illuminated by light from the illumination unit 5, into multiple different polarization components. The light separated into different polarization components is incident on the pixels 33 of the image sensor 32. More specifically, the polarization unit 31 has a plurality of polarizers 34 positioned on the light-receiving surface 35 side of the image sensor 32. Light (reflected light) obtained from the surface to be inspected 2 is incident on the light-receiving surface 35 of the image sensor 32 via the polarization unit 31. The intensity (brightness) of the optical image at the position of each pixel 33 is then detected as a pixel signal. This makes it possible to perform image observation of the surface to be inspected 2.
[0031] Each polarizer 34 is approximately the same size as one of the multiple pixels 33 of the image sensor 32, and is arranged to correspond to the multiple pixels 33 of the image sensor 32. That is, the polarization unit 31 is configured such that one polarizer 34 is placed on the light-receiving surface 35 side of one pixel 33. Therefore, the number of polarizers 34 and the number of pixels 33 are equal.
[0032] Each of the multiple polarizers 34 has a polarization axis with a different polarization direction. For example, when light is incident on a polarizer 34, a polarization component (linear polarization) having a polarization direction parallel to the polarization axis of that polarizer 34 is extracted. In Figure 3(B), the polarizer 34a has a polarization axis in the polarization direction of angle α. The polarization direction of angle α is defined as the reference direction. In this embodiment, angle α is set to 0 degrees. The polarizer 34b has a polarization axis with a polarization direction of angle (α+45) degrees. In other words, it has a polarization axis with an angle of 45 degrees, rotated 45° from the reference direction in a predetermined direction. The polarizer 34c has a polarization axis with a polarization direction of (α+90) degrees. In other words, it has a polarization axis with a 90-degree angle, rotated 90° from the reference direction in a predetermined direction. The polarizer 34d has a polarization axis with a polarization direction of angle (α+135) degrees. In other words, it has a polarization axis with an angle of 135 degrees, rotated 135° from the reference direction in a predetermined direction.
[0033] In the polarization camera 3, light obtained from the inspection surface 2 is incident on the polarization unit 31 (multiple polarizers 34). The multiple polarizers 34 extract polarization components parallel to their respective polarization axes from the incident light and cause the extracted polarization components to be incident on the corresponding pixels 33. It can also be said that the multiple polarizers 34 control the polarization direction of the light traveling toward the corresponding pixels 33.
[0034] Multiple polarizers 34 are formed on the light-receiving surface 35 side of each pixel 33, for example, in accordance with the process of generating multiple pixels 33 of the image sensor 32. That is, a polarization unit 31 is composed of multiple polarizers 34 formed on multiple pixels 33. The specific configuration of the polarizers 34 is not limited, and polarizers 34 using wire grids, liquid crystal elements, polarizing films, etc., may be used as appropriate.
[0035] Thus, the polarization camera 3 shown in Figure 3 can acquire pixel signals for different polarization directions, or in other words, polarization images, from the pixel signals obtained in a single image capture.
[0036] The polarization camera may also have the configuration shown in Figure 4. Figure 4 is a schematic diagram showing another example of polarization camera 3. The polarization camera 3 shown in Figure 4 has an image sensor 37 as an imaging unit and a rotatable polarizing plate 36 as a polarization separation unit. Here, a configuration without a color filter is described, but the polarization camera 3 may have a color filter.
[0037] The image sensor 37 has multiple pixels, each capable of outputting a pixel signal. A CMOS sensor, a CCD sensor, or the like may be used as the image sensor 37 as appropriate. A rotatable polarizing plate 36 is provided on the light-receiving surface side of the polarizing camera 3. By rotating the polarizing plate 36, the light obtained from the inspection surface 2 is separated into multiple polarization components with different polarization directions in a time-division manner. The separated light is incident on the pixels of the image sensor 37, and a pixel signal is output from each pixel. In the polarization camera 3 shown in Figure 4, multiple images are taken while rotating the polarizing plate 36, acquiring multiple polarization images with different polarization directions.
[0038] Furthermore, from the viewpoint of shortening inspection time, it is more preferable to use the polarization camera 3 with the structure shown in Figure 3. In order to acquire pixel signals with multiple different polarization directions, the polarization camera 3 shown in Figure 4 requires rotating the polarizing plate 36 and taking multiple images, whereas the polarization camera 3 shown in Figure 3 only requires one image. Thus, by using the polarization camera 3 shown in Figure 3, inspection time can be shortened and inspection can be performed efficiently.
[0039] In the polarizing camera 3, the image sensor 32 captures an image of the inspection surface through a lens (objective lens and imaging lens) not shown. Polarization camera 3 may be a camera having a Scheinproof optical system. In this technology, in order to observe the change in the state of light due to reflection (change in polarization state), it is preferable to arrange the illumination unit 5 and the polarization camera 3 in a specular reflection configuration, and the illumination unit 5 is positioned so that light from the illumination unit 5 is incident on the inspection surface 2 at an angle. In this case, the focus in the depth direction may be blurred due to the effect of the depth of field of the lens, but by using a camera with a Scheinproof optical system, it is possible to acquire a wide range of high-resolution images in the depth direction. For example, when the field of view is oblique, the depth of field of the camera may cause the foreground and background of the field of view to be out of focus, but this can be avoided with a Scheinproof optical system. A camera with a Shineproof optical system has a movable mechanism that changes the angle between the lens and the horizontal plane, for example, so that the surface to be inspected, the main surface of the lens, and the light-receiving surface of the image sensor 32 satisfy the Shineproof condition.
[0040] (Display) The display unit 6 displays an image of the inspection target surface 2, which has been processed by the processing unit 4 to be suitable for defect inspection. For example, a display device such as an LCD monitor is used as the display unit 6. The display unit 6 is installed near the vehicle body painting line. This allows the inspector to perform defect inspection of the inspection target surface 2 while checking the image displayed on the display unit 6.
[0041] (processing) The processing unit 4 calculates polarization parameters using the pixel signals for each pixel 33 output from the image sensor 32. The processing unit 4 uses these polarization parameters to process the pixel signals so that they become an image suitable for defect inspection of the surface 2 to be inspected.
[0042] In this technology, a polarization phase difference image obtained using polarization parameters is filtered at a predetermined spatial frequency to generate a spatial frequency optimized image. Based on this spatial frequency optimized image, it is possible to accurately detect multiple types of defects on the inspection surface 2 simultaneously from a single image.
[0043] The following describes the processing performed in Processing Unit 4, with reference to Figure 5. Figures 5(A) to (D) are images generated using the pixel signals of the same inspection surface 2. The inspection surface 2 has a curved surface. The inspection surface 2 also has three types of defects: bumps 22, scratches 21, and dents 23. In Figures 5(A), (C), and (D), the wide, linear object located in the upper left is a verification tape used to confirm the location of the dent, and does not represent the original shape or defect of the surface being inspected. "Original shape of the surface being inspected" refers to the shape of the surface being inspected when it is free of defects.
[0044] Figure 5(A) is an image obtained by averaging four polarization images based on the pixel signals of each pixel 33 corresponding to each polarizer 34a to 34d output from the image sensor 32. Figure 5(B) is a polarization degree image that visualizes the distribution of polarization degree (DoP), which is a polarization parameter calculated using the pixel signal for each pixel 33 output from the image sensor 32. Polarization degree will be explained later. Figure 5(C) is a polarization phase difference image that visualizes the distribution of polarization phase difference, which is a polarization parameter calculated using the pixel signal for each pixel 33 output from the image sensor 32. Polarization phase difference will be explained later. Figure 5(D) is a spatial frequency optimized image obtained by filtering the polarization phase difference image of Figure 5(C) at a predetermined spatial frequency. The filtering process will be described later.
[0045] In Figure 5(A), there is a region of varying shades that divides the image into left and right halves. This is because the difference in the inclination of the surface, which is the original shape of the surface to be inspected 2, is reflected as a difference in color intensity. In Figure 5(A), scratches 21 and blemishes 22 are recognizable, but dents 23 are difficult to recognize. This technology allows for the easy detection of multiple types of defects, such as scratches 21, blemishes 22, and dents 23, from a single image, even if the surface to be inspected 2 is curved, by filtering the polarization phase difference image at a predetermined spatial frequency. A detailed explanation follows below.
[0046] The processing unit 4 uses the pixel signal for each pixel 33 output from the image sensor 32 to calculate the Stokes vector S(S0, S1, S2) for each pixel 33 using equation (1) below. The Stokes vector S is a polarization parameter. (1) In equation (1), I0 represents the brightness data (pixel signal) at the pixel 33 corresponding to the polarizer 34a, which has a polarization axis set in the reference direction. 45 This shows the brightness data (pixel signal) in the pixel 33 corresponding to the polarizer 34b, which has a polarization axis set that is rotated 45° from the reference direction in a predetermined direction. 90 This shows the brightness data (pixel signal) in the pixel 33 corresponding to the polarizer 34c, which has a polarization axis set to be rotated 90° from the reference direction in a predetermined direction. 135 This shows the luminance data (pixel signal) in a pixel 33 corresponding to a polarizer 34d whose polarization axis is set to be rotated 135° from the reference direction in a predetermined direction. The luminance data is intensity (brightness) data.
[0047]
number
[0048] The processing unit 4 uses the Stokes vector obtained above to calculate the degree of polarization DoP using equation (2) below. The degree of polarization DoP is a polarization parameter. The processing unit 4 generates a degree of polarization image that visualizes the distribution of the degree of polarization DoP.
[0049]
number
[0050] As shown in Figure 5(B), the polarization degree image, which visualizes the distribution of polarization degree DoP calculated by equation (2) above, clearly shows the scratch 21. Since the light obtained from the scratch 21 mainly consists of scattered reflection components, observing the polarization degree is suitable for detecting the scratch 21. In contrast, it is difficult to detect blemishes 22 and dents 23 by observing the polarization degree.
[0051] The processing unit 4 uses the Stokes vector obtained above to calculate the polarization phase difference Φ using equation (3) below. The polarization phase difference Φ is a polarization parameter. The processing unit 4 generates a phase difference image that visualizes the distribution of the polarization phase difference Φ.
[0052]
number
[0053] As shown in Figure 5(C), the phase difference image, which visualizes the distribution of polarization phase difference Φ calculated by equation (3) above, allows for the recognition of scratches 21, bumps 22, and dents 23. Since the angle of specular reflection changes in light obtained from uneven areas such as scratches 21, blemishes 22, and dents 23, observing the polarization phase difference is suitable for detecting scratches 21, blemishes 22, and dents 23. However, as shown in Figure 5(C), when the surface to be inspected 2 has a curved surface, a color representing the original curved shape of the surface to be inspected 2 (the shaded areas that divide the surface into left and right halves in Figure 5(C)) appears. Therefore, it is difficult to distinguish between a change in the reflection angle of specular reflection caused by the original shape of the surface to be inspected 2 and a change in the reflection angle caused by a defect.
[0054] In contrast, in this technology, the processing unit 4 filters the polarization phase difference image at a predetermined spatial frequency to generate a spatial frequency optimized image. More specifically, a high-pass filter is used to retain the high-spatial-frequency components in the polarization phase difference image while removing the low-spatial-frequency components. This makes it possible to separate the change in the reflection angle of specular reflection caused by the original shape of the inspection surface 2 from the change in the reflection angle caused by defects. The cutoff frequency during filtering can be appropriately set according to the curvature of the surface to be inspected 2. This technology enables defect inspection without changing the relative position of the irradiation unit 5 and the polarization camera 3. Therefore, since the area of the surface to be inspected can be determined in advance, the cutoff frequency can be set in advance according to the curvature of the surface to be inspected.
[0055] As shown in Figure 5(D), in the spatial frequency optimized image obtained by filtering, from the perspective of the original surface 2 being inspected, the boundary between surfaces with different inclinations appears as a roughly straight line extending vertically, dividing the image in half horizontally, because it is a region with a rapid change in brightness, i.e., a region with a high spatial frequency. On the other hand, the parts other than the boundary are regions with a low spatial frequency and therefore do not appear in the image. Thus, the spatial frequency optimized image is an image in which at least some of the information related to the original surface 2 being inspected has been lost. As a result, the spatial frequency optimized image becomes an image in which defects with large local variations, such as scratches 21, blemishes 22, and dents 23, stand out more clearly. Therefore, it becomes possible to recognize defects more clearly, and the defect detection accuracy is improved.
[0056] In this way, by filtering the phase difference image at a predetermined spatial frequency, it becomes easier to distinguish between changes in the reflection angle of specular reflection caused by the original shape of the inspection target surface 2 and changes in the reflection angle caused by defects, such as scratches 21, bumps 22, and dents 23, which are small, locally located areas with uneven surfaces, thereby improving the defect detection accuracy.
[0057] As described above, this technology filters the polarization phase difference image obtained using the pixel signal output from the image sensor 32 at a predetermined spatial frequency. As a result, even if the surface to be inspected 2 has a curved shape, it becomes possible to accurately detect multiple types of defects such as scratches 21, blemishes 22, and dents 23 simultaneously from a single spatial frequency optimized image, thereby shortening inspection time and significantly improving inspection efficiency. Furthermore, in this embodiment, the light-emitting surface of the irradiation unit 5 is made large, which expands the area that can be inspected. Therefore, it becomes possible to detect multiple types of defects over a wide area, and even for large inspection targets such as a vehicle body M, the inspection time can be shortened. While scratches, dents, and bumps were used as examples of defects here, drips and blemishes can also be detected in a similar manner.
[0058] In addition to observing the spatial frequency-optimized image, the polarization degree image (see Figure 5(B)) may also be observed. This can further improve the accuracy of scratch detection.
[0059] This technology is an inspection device suitable for inspecting the surface condition of a target surface. It can inspect multiple types of different defects solely by processing the waveform of the pixel signals obtained from the polarizing camera, without moving the polarizing camera or the illumination unit. Since it does not require setting different lighting conditions for each type of defect as in conventional methods, this technology can reduce the number of optical systems and lights used in the inspection process. Furthermore, by performing defect inspections using spatial frequency-optimized images, it is possible to suppress inspection errors and omissions caused by variations in inspections by different inspectors, compared to conventional visual defect inspections, enabling stable defect inspection work with high inspection accuracy.
[0060] [Testing Method] The inspection method of this technology using the inspection device 1 described above will be explained using the flowchart in Figure 6. As shown in Figure 6, the processing unit 4 acquires the pixel signal for each pixel 33 output from the image sensor 32 of the polarizing camera 3, which captures an image of the inspection target surface 2 illuminated by light from the illumination unit 5 (ST1). Next, the processing unit 4 uses the acquired pixel signal to calculate polarization parameters such as the Stokes vector S, polarization phase difference Φ, and polarization degree DoP (ST2). Next, the processing unit 4 filters the polarization phase difference image based on the calculated polarization phase difference Φ at a predetermined spatial frequency (ST3) to generate a spatial frequency optimized image. The generated spatial frequency optimized image is output to the display unit 6 and displayed on the display unit 6. Defect inspection is performed by the inspector observing the spatial frequency optimized image.
[0061] Furthermore, the processing unit 4 may generate a polarization degree image based on the calculated polarization degree DoP, in addition to the spatial frequency optimized image. The spatial frequency optimized image and the polarization degree image are output to the display unit 6 and displayed on the display unit 6. Defect inspection may be performed by an inspector observing these images.
[0062] The embodiments of this technology are not limited to those described above, and various modifications are possible without departing from the spirit of this technology. For example, in the above embodiment, the surface to be inspected was given as the glossy painted surface of a vehicle body, but it is not limited to this. For example, it can also be applied to surface inspection of any glossy painted surface, resin molded products, semiconductor wafers, transparent parts such as glass and resin, tablets, etc.
[0063] Furthermore, although the above-described embodiment uses the example of a curved surface as the inspection target, this technology can also be applied to the inspection of flat surfaces. However, as mentioned above, this technology can distinguish between changes in the reflection angle of reflected light caused by the original shape of the inspection target and changes in the reflection angle caused by defects through filtering, and is therefore particularly effective for defect inspection of curved surfaces.
[0064] This technology can also take the following configuration: (1) An illumination unit that irradiates the surface to be inspected with light, A polarization separation unit separates the light obtained from the surface to be inspected, which has been irradiated with the aforementioned light, into a plurality of polarization components with different polarization directions. An imaging unit having multiple pixels receives light with multiple different polarization components separated by the polarization separation unit and outputs a pixel signal, A processing unit that filters a polarization phase difference image generated using the pixel signal output from the imaging unit at a predetermined spatial frequency, An inspection device equipped with the following. (2) The inspection device described in (1) above, The surface to be inspected has a curved surface. Inspection device. (3) The inspection device described in (1) or (2) above, The area of the light-emitting surface of the irradiation unit is set to be larger than the surface to be inspected. Inspection device. (4) An inspection device described in any one of (1) to (3) above, The processing unit generates a polarization degree image using the pixel signal output from the imaging unit. Inspection device. (5) An inspection device described in any one of (1) to (4) above, The polarization separation unit has a plurality of polarizers that separate the light obtained from the surface to be inspected into polarization components with different polarization directions, and each of the plurality of polarizers is arranged on the light-receiving surface of the corresponding pixel. Inspection device. (6) The inspection device described in (5) above, Each of the polarizers has polarization axes of angles α, (α+45), (α+90), and (α+135). Inspection device. (7) An inspection device described in any one of (1) to (6) above, The imaging unit has a Scheinproof optical system. Inspection device. (8) The image unit has multiple pixels that receive light from the surface to be inspected that has been irradiated with light, through a polarization separation unit that separates the light into multiple different polarization components, and outputs a pixel signal, and acquires the pixel signal from the image unit. The polarization phase difference image generated using the aforementioned pixel signals is filtered at a predetermined spatial frequency. Testing method. [Explanation of symbols]
[0065] 1…Inspection device 2… Surfaces to be inspected 4… Processing Unit 5…Irradiation area 31…Polarization unit (polarization separation section) 32…Image sensor (imaging unit) 36…Rotating polarizing plate (polarization separation unit) 37…Image sensor (imaging unit)
Claims
1. An irradiation unit that irradiates light onto a curved surface to be inspected, A polarization separation unit separates the light obtained from the surface to be inspected, which has been irradiated with the aforementioned light, into a plurality of polarization components with different polarization directions. An imaging unit having multiple pixels receives light with multiple different polarization components separated by the polarization separation unit and outputs a pixel signal, A processing unit performs a filtering process on a polarization phase difference image generated using the pixel signal output from the imaging unit, using a high-pass filter to remove low spatial frequency components caused by the curved shape of the surface to be inspected, which is set according to the curvature of the surface to be inspected, and to extract high spatial frequency components caused by defects, which are local changes in the surface to be inspected. An inspection device equipped with the following.
2. An inspection apparatus according to claim 1, The area of the light-emitting surface of the irradiation unit is set to be larger than the surface to be inspected. Inspection device.
3. An inspection apparatus according to claim 1, The processing unit generates a polarization degree image using the pixel signal output from the imaging unit. Inspection device.
4. An inspection apparatus according to claim 1, The polarization separation unit has a plurality of polarizers that separate the light obtained from the surface to be inspected into polarization components with different polarization directions, and each of the plurality of polarizers is arranged on the light-receiving surface of the corresponding pixel. Inspection device.
5. The inspection apparatus according to claim 4, Each of the polarizers has polarization axes of angles α, (α+45), (α+90), and (α+135). Inspection device.
6. An inspection apparatus according to claim 1, The imaging unit has a Scheinproof optical system. Inspection device.
7. A polarization separation unit receives light from a curved surface to be inspected that is irradiated with light, separates it into multiple different polarization components, and outputs a pixel signal. The pixel signal is then acquired from an imaging unit having multiple pixels. A high-pass filter is used to perform a filtering process on the polarization phase difference image generated using the aforementioned pixel signals. This filtering process removes low spatial frequency components caused by the curved shape of the surface to be inspected, which is set according to the curvature of the surface to be inspected, and extracts high spatial frequency components caused by defects, which are local changes in the surface to be inspected. Testing method.