Appearance inspection device

The appearance inspection device uses dual light sources and a detection unit to differentiate between abnormal and foreign matter regions, enhancing detection accuracy and reducing false positives.

JP2025114276APending Publication Date: 2025-08-05KK TOYOTA CHUO KENKYUSHO +1
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2024008876
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2024-01-24
Publication Date
2025-08-05

AI Technical Summary

Technical Problem

Existing visual inspection methods falsely detect foreign matter as abnormal areas, leading to decreased yield in objects where foreign matter does not pose a quality problem.

Method used

An appearance inspection device using two light sources with different angles of incidence and a detection unit to distinguish between abnormal regions and foreign matter adhesion regions, excluding overlapping areas from abnormal candidates.

Benefits of technology

Improves detection accuracy by reducing false positives from foreign matter, enhancing yield by accurately identifying true abnormal regions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2025114276000001_ABST
    Figure 2025114276000001_ABST
Patent Text Reader

Abstract

To provide a technique capable of improving the detection accuracy of abnormal regions.SOLUTION: An appearance inspection device for inspecting the appearance of an object comprises a first light source that irradiates light onto the object, a second light source that irradiates light onto the object at a larger incident angle than the first light source, an imaging unit that is disposed at a position where the regular reflection component of the reflected light from the object when the light is irradiated onto the object by the second light source does not enter, and acquires a first image showing the surface of the object when light from the first light source is irradiated and a second image showing the surface of the object when light from the second light source is irradiated, and a detection unit that detects abnormal candidate regions on the surface of the object using the first image, detects foreign matter adhesion regions on the surface using the second image, and then excludes, from the abnormal candidate regions, regions overlapping with the foreign matter adhesion regions.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical Field]

[0001] The present invention relates to a visual inspection apparatus. [Background technology]

[0002] Conventionally, there has been known a technique for inspecting the appearance of an object by capturing an image of the object when it is irradiated with light. For example, Patent Document 1 discloses an inspection system including an inspection illumination device and an imaging device, in which a first light-shielding mask, a second light-shielding mask, a lens, and a half mirror are arranged along the direction in which light irradiated from a surface light source travels. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-180621 Summary of the Invention [Problem to be solved by the invention]

[0004] In visual inspections using images of an object captured under light irradiation, while areas of the object's surface containing scratches, dents, etc. are detected as abnormal areas, areas of the object's surface containing dust or other foreign matter may also be detected as abnormal areas. However, for objects where the presence of dust or other foreign matter does not pose a quality problem, this detection is considered a false positive, and even objects without scratches, dents, etc. may be deemed defective due to the presence of foreign matter, leading to a decrease in yield. Therefore, there has been a demand for technology that can improve the accuracy of detecting abnormal areas in visual inspections of objects by reducing the possibility of falsely detecting areas containing dust or other foreign matter as abnormal areas. However, Patent Document 1 does not consider reducing the possibility of falsely detecting abnormal areas.

[0005] The present invention has been made to solve at least part of the above-mentioned problems, and has an object to provide a technique that can improve the accuracy of detecting an abnormal region. [Means for solving the problem]

[0006] The present invention has been made to solve at least part of the above-mentioned problems, and can be realized in the following aspects.

[0007] (1) According to one aspect of the present invention, there is provided an appearance inspection device for inspecting the appearance of an object, the appearance inspection device including: a first light source that irradiates the object with light; a second light source that irradiates the object with light at a larger angle of incidence than the first light source; an imaging unit that is positioned at a position where a specular reflection component of light reflected from the object when the second light source irradiates the object with light is not incident, and that acquires a first image showing the surface of the object when irradiated with light from the first light source and a second image showing the surface of the object when irradiated with light from the second light source; and a detection unit that detects an abnormal candidate region on the surface of the object using the first image and a foreign matter adhesion region on the surface using the second image, and then excludes a region overlapping the foreign matter adhesion region from the abnormal candidate region.

[0008] According to this configuration, it is possible to exclude from the abnormal candidate region detected using the first image any region that overlaps with a foreign substance-adhered region detected using the second image. That is, because the region that overlaps with a foreign substance-adhered region is excluded from the abnormal candidate region that is a candidate for the abnormal region, it is possible to reduce the possibility of erroneously detecting a region with foreign substance such as dust as an abnormal region. As a result, the detection accuracy of the abnormal region is improved, thereby improving the yield.

[0009] (2) In the appearance inspection device of the above form, the first light source can irradiate light onto the object at different illumination outer diameters, and the detection unit uses a plurality of the first images acquired when light is irradiated from the first light source at different illumination outer diameters to calculate a brightness accumulated value by accumulating the brightness value of each pixel in each of the first images at the same position, and then detects the abnormal candidate area from a brightness accumulated image created by arranging each of the brightness accumulated values corresponding to each position, and the brightness accumulated value may be a value calculated by normalizing the brightness value of each pixel in each of the first images and then accumulating the normalized brightness values at the same position. When an object is covered with machining oil, the intensity of light reflected from the object decreases, and the brightness values of the constituent pixels constituting each position of the object also tend to decrease in an image captured of the object. According to this configuration, a brightness-integrated image created using multiple first images is used to detect an abnormal candidate region. Furthermore, because the brightness values of the constituent pixels constituting each position of the object in the brightness-integrated image are corrected by normalization, even if the object is covered with machining oil or has machining oil partially attached to it, the influence of such machining oil can be reduced and the abnormal candidate region can be accurately detected.

[0010] (3) In the visual inspection device of the above aspect, the second light source may be a ring light that irradiates light from the periphery of the object. If the second light source is not a ring light but a light that irradiates the object from one side, depending on the shape of the object, light may not be irradiated onto a portion of the object's surface, and even if a foreign object is attached to the object's surface that is not irradiated with light, it may not be captured in the second image. On the other hand, with this configuration, when the second image is captured, the object is imaged with light irradiated from all around the object, so light can be irradiated onto the entire surface of the object. As a result, the accuracy of detecting a foreign object-adhered area using the second image can be improved, and ultimately the accuracy of detecting an abnormal area can be further improved.

[0011] The present invention can be realized in various forms, for example, in the form of an appearance inspection device, an appearance inspection system, an appearance inspection method, a method for controlling appearance inspection operations, a computer program for executing these devices and methods, a server device for distributing this computer program, a non-transitory storage medium on which a computer program is stored, etc. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is an explanatory diagram illustrating the configuration of a visual inspection device 1 according to a first embodiment. [Figure 2] 3A to 3C are explanatory diagrams showing cross sections of a first light source. [Figure 3] FIG. 1 is an explanatory diagram for explaining fluctuations in reflected light from the surface of an object. [Figure 4] FIG. 2 is an explanatory diagram for explaining the arrangement of cameras. [Figure 5] FIG. 10 is an explanatory diagram of a method for determining whether or not an object has an abnormality in its appearance. [Figure 6] FIG. 10 is an explanatory diagram showing light irradiation by a second light source of a comparative example. [Figure 7] 4 is an explanatory diagram showing light emitted by a second light source according to the first embodiment. FIG. DETAILED DESCRIPTION OF THE INVENTION

[0013] First Embodiment FIG. 1 is an explanatory diagram illustrating the configuration of a visual inspection apparatus 1 according to a first embodiment of the present invention. FIG. 1 illustrates mutually orthogonal X, Y, and Z axes. The X axis corresponds to the direction of light emitted from a first light source 5 (described later), and the Y and Z axes correspond to directions perpendicular to the X axis. These X, Y, and Z axes are common to FIG. 1 and subsequent figures. The visual inspection apparatus 1 is an apparatus for inspecting the visual appearance of an object OB. Specifically, the visual inspection apparatus 1 is an apparatus for detecting the presence or absence of abnormalities, such as scratches or dents, on the surface SF of the object OB. The visual inspection apparatus 1 includes a first light source 5, a plano-convex lens 40, a half mirror 50, a second light source 60, a camera 70, and a control unit 80. In FIG. 1, an optical axis AX indicates the optical axis of the first light source 5.

[0014] FIG. 2A shows a cross section of the first light source 5 taken along line F2A-F2A in FIG. 1. FIG. 2B shows a cross section of the first light source 5 taken along line F2B-F2B in FIG. 2A. As shown in FIG. 1, the first light source 5 irradiates light onto an object OB. FIG. 1 shows a state in which the object OB is disposed at the disposition position PL. The first light source 5 is a unit in which a plate-shaped member 10, a wall member 20, and a diffusion member 30 are stacked in the X-axis direction. As shown in FIG. 2B, light sources 11-15 that irradiate light onto the object OB are disposed on the surface of the plate-shaped member 10 facing the +X-axis direction. In detail, as shown in FIG. 2A, when viewed from the +X-axis direction, one light source 11 is disposed in the center, and multiple light sources 12 are disposed in a circle surrounding the light source 11. Furthermore, multiple light sources 13 are disposed in a circle surrounding the multiple light sources 12. Similarly, the multiple light sources 14 are arranged in a circle surrounding the multiple light sources 13, and the multiple light sources 15 are arranged in a circle surrounding the multiple light sources 14. That is, the light sources 12 to 15 are arranged concentrically on the surface of the plate-like member 10 facing the +X-axis direction. The light sources 11 to 15 are LEDs. The first light source 5 can irradiate light onto the object OB with different illumination outer diameters by adjusting which of the light sources 11 to 15 emits light. As examples of different illumination outer diameters, FIG. 2(A) shows an illumination outer diameter Da when light is emitted from the light sources 11 to 13 and an illumination outer diameter Db when light is emitted from the light sources 11 to 15.

[0015] As shown in FIG. 2(A), the wall member 20 includes wall portions 21-25 that concentrically surround each of the light sources 11-15 when viewed from the +X-axis direction. Each of the wall portions 21-25 is formed to surround the light sources 11-15 when viewed from the +X-axis direction, and extends along the X-axis direction (see FIG. 2(B)). In detail, each of the wall portions 21-24 is formed in a circular shape when viewed from the +X-axis direction, and is arranged to surround the light sources 11-14. The wall portion 25 is formed in a rectangular parallelepiped shape having a cylindrical through-hole (a space in which the light sources 11-15 and the wall portions 21-24 are arranged) in the center when viewed from the +X-axis direction, and is arranged to surround the light source 15. The diffusion member 30 diffuses the light emitted from the light sources 11-15. In addition, in FIGS. 2(A) and 2(B), the wall portions 21 to 24 are cross-sectional views and should be shown by hatching, but for convenience of illustration, hatching has been omitted.

[0016] Returning to the description of FIG. 1 , the plano-convex lens 40 is disposed between the first light source 5 and a half mirror 50 (described later) in the X-axis direction. The plano-convex lens 40 condenses light emitted from the light sources 11 to 15 via the diffusing member 30. The half mirror 50 is disposed between the first light source 5 and the arrangement position PL. Specifically, the half mirror 50 is disposed between the first light source 5 and the arrangement position PL, on the +X-axis direction side of the plano-convex lens 40. When light is emitted from the light sources 11 to 15 (first light source 5) via the diffusing member 30, the half mirror 50 transmits a portion of the light toward the object OB and reflects a portion of the light reflected from the object OB toward a camera 70 (described later). Furthermore, when light is emitted from a second light source 60 (described later) toward the object OB, the half mirror 50 reflects a portion of the light reflected from the object OB toward a camera 70 (described later).

[0017] The second light source 60 irradiates the object OB with light at a larger angle of incidence than the first light source 5. The angle of incidence here refers to the angle between a light ray incident on the surface of the object OB and a normal to the surface SF of the object OB. In other words, the angle between the normal and the direction in which the light irradiated from the second light source 60 travels toward the object OB is larger than the angle between the normal and the direction in which the light irradiated from the first light source 5 travels toward the object OB. In this embodiment, the second light source 60 is a ring light that irradiates light from the periphery of the object OB.

[0018] The camera 70 will be described later. The control unit 80 is a computer including a ROM, a RAM, and a CPU, and performs various controls of the appearance inspection apparatus 1. The control unit 80 changes the illumination angle θ of the light irradiated onto the object OB by controlling the illumination outer diameter (illustrated as illumination outer diameters Da and Db in FIG. 2(A)) defined by the light irradiated from each of the light sources 11 to 15 (see FIG. 1). For example, the illumination outer diameter is larger when light is irradiated from the light sources 11 to 13 than when light is irradiated only from the light source 11. Similarly, the illumination outer diameter is larger when light is irradiated from the light sources 11 to 15 than when light is irradiated from the light sources 11 to 13. The illumination angle θ increases as the illumination outer diameter increases. The control unit 80 also controls the execution of imaging by the camera 70.

[0019] The camera 70 is an imaging unit that captures light incident therein via the half mirror 50. A telecentric lens 72 attached to the camera 70 allows only light parallel to the optical axis of the telecentric lens 72 to enter the camera 70. The camera 70 is an imaging unit that can capture light that enters the camera 70 via the half mirror 50 when light is being irradiated from the first light source 5 or the second light source 60 toward the object OB placed at the placement position PL. Of the images captured and acquired by the camera 70, an image showing the surface SF of the object OB when light is being irradiated from the first light source 5 is referred to as a first image G1, and an image showing the surface SF of the object OB when light is being irradiated from the second light source 60 is referred to as a second image G2.

[0020] FIG. 3 is an explanatory diagram illustrating how reflected light from the surface SF of the object OB changes depending on the presence or absence of foreign matter FR, such as dust. In the explanation using FIG. 3, it is assumed that the surface SF has perfect specular reflection characteristics and the foreign matter FR has perfect diffuse reflection characteristics. Light In incident on the surface SF is reflected as specular reflection light Re at locations on the surface SF where no foreign matter FR is attached, and is reflected as diffuse reflection light Df at locations on the surface SF where foreign matter FR is attached. In the first image G1 described above, abnormal areas where scratches, dents, etc. are present and foreign matter-attached areas where dust or other foreign matter FR is attached tend to have similar brightness values, making it difficult to distinguish between abnormal areas and foreign matter-attached areas from the first image G1.

[0021] 4 is an explanatory diagram for explaining the arrangement of the camera 70. As explained in FIG. 3, light In incident on the surface SF is reflected as specularly reflected light Re at positions where there are no foreign matter FR and as diffusely reflected light Df at positions where there are foreign matter FR attached. Therefore, by placing the camera 70 at a position where the specularly reflected light Re is not incident and having the camera 70 capture a portion of the diffusely reflected light Df (shown as diffusely reflected light Dp in FIG. 4), it is possible to identify the area where foreign matter is attached. Note that, since the direction of the specularly reflected light Re may change depending on the surface texture and shape of the actual object OB, it is preferable to arrange the light source and the camera 70 so that the angle α (shown in FIG. 4) between the direction of the light In incident on the object OB and the direction in which the camera 70 is positioned relative to the object OB is as close to a right angle as possible. In the appearance inspection device 1 shown in FIG. 1, the camera 70 can be considered to be disposed at a position facing the object OB. Therefore, if a light source is disposed in a direction nearly perpendicular to the direction in which the camera 70 and the object OB face each other (the X-axis direction in FIG. 1), specularly reflected light Re emitted from the light source and then reflected from the surface SF can be made less likely to enter the camera 70. For this reason, in the appearance inspection device 1 shown in FIG. 1, the second light source 60 is disposed at a position closer to the -X-axis direction than the object OB (disposition position PL) and at a position where light can be incident on the surface SF of the object OB from a direction that forms an angle with the X-axis direction nearly perpendicular. In other words, the camera 70 is disposed at a position where the specularly reflected component of the light reflected from the object OB when light is irradiated onto the object OB from the second light source 60 does not enter. Therefore, in the second image G2 showing the surface SF of the object OB when light is irradiated from the second light source 60, the luminance values of areas where scratches, dents, etc. are present and areas where no foreign matter FR is attached are relatively small, and the luminance values of areas where foreign matter FR is attached are relatively large. Note that in the first image G1 showing the surface SF of the object OB when light is irradiated from the first light source 5, the luminance values of areas of the surface that are closer to being perpendicular to the optical axis AX are relatively large, and the luminance values of areas where foreign matter FR is attached and areas where scratches, dents, etc. are present are relatively small.

[0022] FIG. 5 is an explanatory diagram of a method for determining whether or not there is an abnormality in the appearance of the object OB. In the appearance inspection device 1, the control unit 80 functions as a detection unit, detecting the presence or absence of an abnormality, such as a scratch or a dent, on the surface SF of the object OB using the first image G1 and the second image G2 acquired by the camera 70. The details of this method are described below. First, the control unit 80 creates a brightness-integrated image Li using an image group IG consisting of multiple first images G1 acquired when light is irradiated from the first light source 5 at different illumination outer diameters. In detail, the control unit 80 calculates a brightness-integrated value by integrating the brightness value of each pixel in each of the first images G1 included in the image group IG at the same position, and then creates the brightness-integrated image Li by arranging each of the calculated brightness-integrated values corresponding to each position.

[0023] The above-mentioned multiple first images G1 refer to multiple images captured each time the illumination outer diameter of the light emitted from the first light source 5 is changed to a different size. For example, if the camera 70 captures images in each of the following states: when light is emitted only from light source 11; when light is emitted from light sources 11-13; and when light is emitted from light sources 11-15, three images correspond to the above-mentioned multiple images. Furthermore, calculating the integrated brightness value by integrating the brightness values of each pixel at the same position in each of the first images G1 means, for example, that if the above-mentioned multiple first images G1 are three images, there are three pixels at the same position in each image, and the three pieces of data are integrated at the same position to calculate the integrated brightness value. Arranging each of the integrated brightness values corresponding to each position means, for example, assuming that the integrated brightness value integrated for coordinates (Xi, Yj) corresponds to coordinates (Xi, Yj), and arranging the integrated brightness value at the position of coordinates (Xi, Yj). That is, the luminance value of each pixel constituting the luminance accumulated image Li is a luminance accumulated value calculated based on a plurality of first images G1.

[0024] In this embodiment, the integrated brightness value is a value calculated by normalizing the brightness value of each pixel in each first image G1 and then integrating the normalized brightness values for each corresponding position. Normalization refers to a process in which the difference between the brightness value of each pixel constituting the same position in each first image G1 and the minimum brightness value of the pixel constituting the same position is divided by the difference between the maximum and minimum brightness values of the pixel constituting the same position. Through this normalization, the brightness value of each pixel constituting the same position is converted into a value within a range of 0 to 1. When the object OB is covered with machining oil, the intensity of light reflected from the object OB decreases, which tends to decrease the brightness value of the pixel constituting each position of the object OB in an image captured of the object OB. In response to this, normalizing the brightness values can reduce the effects of machining oil.

[0025] Returning to the description of the method for determining whether or not there is an abnormality in the appearance of the object OB, as shown in Fig. 5, the control unit 80 creates a brightness integrated image Li, and then obtains a difference image Di between a reference image ST (described later) and the brightness integrated image Li. The reference image ST is a brightness integrated image created from a group of images of an object OB having no abnormalities in appearance. The difference image Di is obtained by taking the difference in brightness values between pixels that constitute the same position of the object OB in each of the reference image ST and the brightness integrated image Li.

[0026] Next, the control unit 80 calculates the average value and standard deviation σ of the integrated brightness values of each pixel constituting the differential image Di, and then, as shown in the graph GR in FIG. 5, determines pixels having integrated brightness values within the average value ±kσ (k is an arbitrary constant) as normal portions NM, and determines pixels having integrated brightness values not within the average value ±kσ as abnormal candidate portions AC. The graph GR is a histogram of the integrated brightness values of each pixel constituting the differential image Di. Next, the control unit 80 obtains an extracted image Ex1 created by extracting only pixels determined to be abnormal candidate portions AC from among the pixels constituting the differential image Di. The extracted image Ex1 in FIG. 7 shows abnormal candidate regions Aa1-Aa3 composed of pixels determined to be abnormal candidate portions AC.

[0027] Next, the control unit 80 obtains an extracted image Ex3 by excluding areas that overlap with the foreign substance adhesion area Fa from the abnormal candidate areas Aa1-Aa3 included in the extracted image Ex1. Specifically, the control unit 80 obtains an extracted image Ex2 by applying binarization to the second image G2 captured by the camera 70 (an image showing the object OB when light is irradiated from the second light source 60), and then obtains the extracted image Ex3 by excluding from the extracted image Ex1 areas that overlap with the foreign substance adhesion area Fa, which are shown as areas with high brightness values in the extracted image Ex2. The area that overlaps with the foreign substance adhesion area Fa refers to the area in the extracted image Ex1 that overlaps with the foreign substance adhesion area Fa when the extracted images Ex1 and Ex2 are superimposed. In the extracted image Ex3 of FIG. 7, the abnormal candidate area Aa3, of the abnormal candidate areas Aa1-Aa3, is excluded from the abnormal candidate area because it overlaps with the foreign substance adhesion area Fa. In other words, it is highly likely that the abnormal candidate region Aa3 is not an abnormal region where scratches, dents, etc. are present, but rather a foreign matter adhesion region where a foreign matter FR is adhered. In this way, the control unit 80 detects the abnormal candidate region (corresponding to the abnormal candidate regions Aa1 to Aa3 in FIG. 7) on the surface SF of the object OB using the first image G1, and detects the foreign matter adhesion region (corresponding to the foreign matter adhesion region Fa in FIG. 7) on the surface SF of the object OB using the second image G2, and then functions as a detection unit that excludes regions that overlap with the foreign matter adhesion region from the abnormal candidate region. Note that in this embodiment, the control unit 80 detects the abnormal candidate region from the brightness integrated image Li created using multiple first images G1. Here, in FIG. 7, the entire abnormal candidate region Aa3 is excluded from the abnormal candidate region as a region that overlaps with the foreign matter adhesion region Fa, but of course, the area of the excluded abnormal candidate region varies depending on the degree of overlap with the foreign matter adhesion region Fa. That is, as shown in extracted images Ex1 and Ex3 in FIG. 7, in addition to cases where an entire abnormal candidate region is excluded, there are also cases where only a part of an abnormal candidate region is excluded.

[0028] Next, the control unit 80 draws a rectangle circumscribing the abnormal candidate regions Aa1 and Aa2 in the extracted image Ex3. The control unit 80 then determines whether the short side SS of this rectangle is equal to or greater than a preset length, and detects, as an abnormal candidate region, any abnormal candidate region circumscribing a rectangle defined by a short side SS equal to or greater than the preset length. On the other hand, the control unit 80 does not detect, as an abnormal region, any abnormal candidate region circumscribing a rectangle defined by a short side SS less than the preset length.

[0029] As described above, the visual inspection apparatus 1 of the first embodiment can exclude from the abnormal candidate region detected using the first image G1 an area that overlaps with a foreign substance-adhered area detected using the second image G2. That is, because the area that overlaps with a foreign substance-adhered area is excluded from the abnormal candidate region that is a candidate for the abnormal region, it is possible to reduce the possibility of erroneously detecting an area with dust or other foreign substance attached as an abnormal region. As a result, the detection accuracy of the abnormal region is improved, thereby improving the yield.

[0030] As described above, in the first image G1, the brightness value of a surface region closer to perpendicular to the optical axis AX is relatively higher, while the brightness values of regions with foreign matter FR and regions with scratches, dents, etc. are relatively lower. Therefore, when only the first image G1 is used, it is difficult to distinguish between an abnormal region where scratches, dents, etc. are present and a foreign matter-adhered region where dust or other foreign matter FR is attached, and there is a risk of the foreign matter-adhered region being erroneously detected as an abnormal region. In this regard, the appearance inspection apparatus 1 of the first embodiment can distinguish between an abnormal region and a foreign matter-adhered region by using the first image G1 and the second image G2 and excluding regions that overlap with the foreign matter-adhered region from abnormal candidate regions that correspond to provisional abnormal regions.

[0031] When the object OB is covered with machining oil, the intensity of light reflected from the object OB decreases, and the brightness values of the constituent pixels constituting each position of the object OB also tend to decrease in an image captured of the object OB. In this regard, in the appearance inspection device 1 of the first embodiment, a brightness integrated image Li created using multiple first images G1 is used to detect an abnormal candidate region. Furthermore, because the brightness values of the constituent pixels constituting each position of the object in the brightness integrated image Li are corrected by normalization, even if the object OB is covered with machining oil or has machining oil partially attached to the object OB, the influence of such machining oil can be reduced and an abnormal candidate region can be accurately detected.

[0032] 6 is an explanatory diagram showing light irradiation by the second light source 60 in a comparative example. When the second light source 60 is not a ring light but an illuminator that irradiates the object OB with light Ic from one side, depending on the shape of the object OB (for example, the shape of the object Ob shown in FIG. 6), the light Ic may not be irradiated onto a part of the surface SF of the object OB, and even if a foreign matter FR is attached to the part of the surface SF of the object OB that is not irradiated with light Ic, it may not be captured in the second image G2.

[0033] FIG. 7 is an explanatory diagram showing the illumination of light by the second light source 60 in the first embodiment. Compared to the comparative example described in FIG. 6, in the visual inspection device 1 of the first embodiment, the second light source 60 is a ring light that illuminates light IC from the periphery of the object OB. Therefore, since light IC is illuminated from the periphery of the object OB, the second image G2 can be captured in a state in which light IC is illuminated on the entire surface SF of the object OB. As a result, the accuracy of detecting a foreign matter adhesion region using the second image G2 can be improved, and ultimately, the accuracy of detecting an abnormal region can be further improved. Similar to FIG. 4, FIG. 7 shows diffuse reflection light Dp that is incident on the camera 70 out of the diffuse reflection light Df reflected from the position where the foreign matter FR is attached.

[0034] <Modification of this embodiment> The present invention is not limited to the above-described embodiment, and can be embodied in various forms without departing from the spirit of the invention. For example, the following modifications are also possible.

[0035] In the above-described embodiment, the light sources 11 to 15 included in the first light source 5 are LEDs, but this is not limiting. The light sources 11 to 15 included in the first light source 5 may be point light sources other than LEDs (for example, halogen lamps, etc.). Furthermore, in the above-described embodiment, the size of the illumination outer diameter is changed by adjusting which of the light sources 11 to 15 is to emit light as the first light source 5, but this is not limiting. For example, a liquid crystal display that can emit light all or part of the screen may be used as the first light source 5, and the size of the illumination outer diameter may be changed by specifying an area to be emitted on the liquid crystal display, thereby changing the illumination outer diameter.

[0036] In the above-described embodiment, the second light source 60 is a ring light that irradiates light from the periphery of the object OB, but this is not limited to this. The second light source 60 may be an illuminator that irradiates light onto the object OB from one side. Of course, from the viewpoint of improving the detection accuracy of the foreign matter adhesion region, it is preferable that the second light source 60 is a ring light.

[0037] In the above-described embodiment, the abnormal candidate region is detected from the brightness integrated image Li, but this is not limiting. For example, the abnormal candidate region may be detected from a single first image G1. That is, the single first image G1 does not have to be an image created from multiple first images G1, as in the brightness integrated image Li, but may be a single image captured when light is irradiated from the first light source 5 with a certain illumination outer diameter, and may be an image that has not been processed using data from other images. In this case, the difference image Di is acquired using the single first image G1 and a single reference image captured of an object OB with a normal appearance.

[0038] In the above-described embodiment, the luminance values of each pixel constituting the same position of the object OB in the multiple first images G1 are converted into values within a range of 0 to 1 after normalization, but this is not limited to this. The maximum value of the range of luminance values after normalization may be reassigned a numerical value other than 1, and the minimum value may be reassigned a numerical value other than 0. For example, after normalization, the maximum value may be reassigned to 100 and the minimum value may be reassigned to -100. Alternatively, the minimum value may remain 0 and only the maximum value may be reassigned to 255. In other words, any values may be reassigned to the maximum and minimum values as long as the distribution of luminance values of each pixel included in the range from the minimum value to the maximum value does not change from before normalization.

[0039] This aspect has been described above based on embodiments and modifications. However, the above-described embodiments are intended to facilitate understanding of this aspect and are not intended to limit this aspect. This aspect may be modified or improved without departing from the spirit and scope of the claims, and equivalents thereof are included in this aspect. Furthermore, if a technical feature is not described as essential in this specification, it may be deleted as appropriate. [Explanation of symbols]

[0040] 1...Visual inspection device 5...1st light source 10...Plate-shaped member 11~15…Light source 20...Wall components 21~25…Wall part 30...Diffusion element 40...Plano-convex lens 50...Half mirror 60…Second light source 70...Camera 72...Telecentric lens 80...Control unit

Claims

1. An appearance inspection device that inspects the appearance of an object, a first light source that irradiates the object with light; a second light source that irradiates the object with light at a larger angle of incidence than the first light source; an imaging unit that is disposed at a position where a specular reflection component of light reflected from the object when the object is irradiated with light from the second light source does not enter, and that acquires a first image showing the surface of the object when the object is irradiated with light from the first light source and a second image showing the surface of the object when the object is irradiated with light from the second light source; a detection unit that detects an abnormal candidate region on the surface of the object using the first image and detects a foreign matter adhesion region on the surface using the second image, and then excludes a region that overlaps with the foreign matter adhesion region from the abnormal candidate region.

2. 2. The visual inspection apparatus according to claim 1, the first light source is capable of irradiating the object with light at different illumination outer diameters; the detection unit uses a plurality of the first images acquired when light is irradiated from the first light source at different illumination outer diameters to calculate a brightness integrated value by integrating the brightness values of each pixel in each of the first images at the same position, and then detects the abnormal candidate region from a brightness integrated image created by arranging each of the brightness integrated values corresponding to each position; An appearance inspection apparatus, wherein the brightness integrated value is a value calculated by normalizing the brightness value of each pixel in each of the first images and then integrating the normalized brightness values at the same position.

3. 3. The visual inspection apparatus according to claim 1, The second light source is a ring light that irradiates light from around the object.

Citation Information

Patent Citations

  • Lighting device for inspection, and inspection system

    JP2016180621A