JP2020079755A - Image inspection apparatus - Google Patents

Abstract

To obtain an image inspection apparatus that more easily detects abnormality.SOLUTION: An image inspection apparatus according to the present disclosure comprises: an image acquisition section that acquires a first image being imaged in a state, in which an inspection object surface being a surface of a long-sized inspection object is irradiated with light from a first directional back side along a longitudinal direction of the inspection object, and a second image being imaged in a state, in which the inspection object surface is irradiated with light from a front side in a first direction; a shading compensation section that acquires a first shading image which is shading-compensated from the first image and that acquires a second shading image which is shading-compensated from the second image; and an abnormality detection section that detects abnormality from a difference image between the first shading image and the second shading image.SELECTED DRAWING: Figure 7

Description

  The present disclosure relates to an image inspection device.

  2. Description of the Related Art Conventionally, there is known an inspection device that detects an abnormality in an inspection target object from an image of a long inspection target object (for example, Patent Document 1).

JP, 2011-011496, A

  In an inspection apparatus that detects an abnormality from an image, for example, if the image has uneven brightness, it is difficult to detect the abnormality.

  Therefore, an object of the present disclosure is, for example, to obtain an image inspection device in which it is easier to detect an abnormality.

  In the image inspection apparatus of the present disclosure, the inspection target surface, which is the surface of the elongated inspection target, is photographed in a state in which light is irradiated from the rear side in the first direction along the longitudinal direction of the inspection target. A first image, an image acquisition unit that acquires a second image captured in a state where the surface to be inspected is irradiated with light from the front side in the first direction, and a shading-corrected first image from the first image. Abnormality that detects an abnormality from a difference image between the first shading image and the second shading image, and a shading correction unit that obtains one shading image and also obtains a second shading corrected image from the second image And a detection unit.

  According to the present invention, for example, it is possible to obtain an image inspection apparatus in which it is easier to detect an abnormality.

FIG. 1 is an exemplary and schematic perspective view of an inspection device of an embodiment. FIG. 2 is an exemplary schematic view of a part of the inspection apparatus according to the embodiment as viewed from the longitudinal direction of the inspection object. FIG. 3 is an explanatory diagram of the first image and the second image acquired by the inspection device of the embodiment. FIG. 4 is an exemplary schematic diagram of a first image acquired by the inspection device of the embodiment. FIG. 5 is an exemplary block diagram of the inspection apparatus of the embodiment. FIG. 6 is an exemplary block diagram of the control unit of the inspection apparatus of the embodiment. FIG. 7 is an exemplary block diagram of the image processing unit of the inspection apparatus of the embodiment. FIG. 8 is an exemplary flowchart of an abnormality detection procedure performed by the inspection device of the embodiment. FIG. 9 is an exemplary schematic view of a part of the first image acquired by the inspection device of the embodiment. FIG. 10 is an exemplary schematic view of a part of the second image acquired by the inspection device of the embodiment. FIG. 11 is an exemplary schematic view of the first shading image in the inspection device of the embodiment. FIG. 12 is an exemplary schematic diagram of the second shading image in the inspection device of the embodiment. FIG. 13 is an exemplary schematic diagram of a difference image between the first shading image of FIG. 11 and the second shading image of FIG. 12 in the inspection device of the embodiment. FIG. 14 is an exemplary schematic view of a first shading image different from that in FIG. 11 in the inspection device of the embodiment. FIG. 15 is an exemplary schematic view of a second shading image different from that in FIG. 12 in the inspection device of the embodiment. FIG. 16 is an exemplary schematic diagram of a difference image between the first shading image of FIG. 14 and the second shading image of FIG. 15 in the inspection device of the embodiment. FIG. 17 is an exemplary plot diagram showing the distribution of the first index and the second index of the plurality of abnormality candidate regions in the inspection device of the embodiment.

  Hereinafter, exemplary embodiments of the present invention will be disclosed. The configurations of the embodiments shown below, and the actions and results (effects) provided by the configurations are examples. The present invention can be realized by a configuration other than the configurations disclosed in the following embodiments. Further, according to the present invention, at least one of various effects (including derivative effects) obtained by the disclosed configuration and a similar configuration can be obtained. Further, in the present specification, the ordinal numbers are given for convenience of distinguishing parts, parts, and the like, and do not indicate the priority order or the order.

  FIG. 1 is a perspective view of the inspection device 1. The inspection apparatus 1 shown in FIG. 1 inspects the surface 100a of the inspection object 100 based on the image of the inspection object 100. The inspection object 100 is, for example, a long and cylindrical or columnar component such as a hose, a tube, or a rod. The surface 100a is an example of a surface to be inspected.

  As shown in FIG. 1, the inspection device 1 includes a plurality of light sources 2U and 2D. Specifically, the light source 2U and the light source 2D are arranged apart from each other in the longitudinal direction of the inspection object 100. The light source 2U and the light source 2D are configured in an annular shape, for example, an annular shape. The light source 2U and the light source 2D have a plurality of light emitted diodes (LEDs) arranged in an annular shape. The elongated inspection object 100 penetrates through the rings of the light sources 2U and 2D and extends along the axial direction of the annular portion. Further, the light source 2U and the light source 2D are arranged in plane symmetry with respect to a plane orthogonal to the longitudinal direction of the inspection object 100. The longitudinal direction of the inspection object 100 may also be referred to as an axial direction.

  The imaging region A of the inspection object 100 is set between the light source 2U and the light source 2D, for example, at an intermediate position between the light source 2U and the light source 2D. Light from the light source 2U is applied to the imaging region A from one side in the longitudinal direction of the inspection object 100, that is, from the left side in FIG. 1, and light from the light source 2D is from the other side in the longitudinal direction, from the right side in FIG. The imaging area A is irradiated.

  The inspection object 100 is conveyed in the longitudinal direction (axial direction) by the conveying device 30 (see FIG. 5) during the inspection. That is, the inspection device 1 inspects the inspection object 100 which is conveyed and moved by the conveyance device 30. Therefore, the imaging region A moves along the longitudinal direction of the inspection target 100 as the inspection target 100 moves. The inspection object 100 moves from the light source 2U side to the light source 2D side. That is, the light source 2U is located behind the imaging area A in the carrying direction, and the light source 2D is located ahead of the imaging area A in the carrying direction.

  The imaging unit 3 (3U, 3D) is located on the outer side in the radial direction of the inspection object 100, and acquires an image of the surface 100a (outer surface, peripheral surface, side surface) of the inspection object 100. That is, the imaging region A is a part of the surface 100a of the inspection object 100.

  FIG. 2 is a view of a part of the inspection device 1 viewed from the longitudinal direction of the inspection object 100. The imaging unit 3 acquires an image of the imaging area A via the optical component 4. The optical component 4 is a plurality of mirrors 4L and 4R. As shown in FIG. 2, the mirrors 4L and 4R are arranged in a V shape on the side opposite to the imaging unit 3 of the inspection object 100 as viewed from the axial direction. The mirrors 4L and 4R are arranged at an angle of 120°, in other words, at an angle of 60° on the opposite side with respect to the line L connecting the imaging unit 3 and the inspection object 100. Each of the mirrors 4L and 4R has a planar reflecting surface 4a facing the surface 100a of the inspection object 100. The reflecting surface 4a extends along a surface substantially orthogonal to the radial direction of the inspection object 100. The mirrors 4L and 4R extend in a strip shape along a plane orthogonal to the longitudinal direction of the inspection object 100. In such a configuration, the imaging region A is a region of the front surface 100a of the inspection object 100 for approximately a half circumference facing the mirrors 4L and 4R. Although the adjacent mirrors 4L and 4R are integrated, they may be separated.

  The inspection units 10 (10U, 10D) including the light sources 2U, 2D, the imaging unit 3, and the optical component 4 are provided at a plurality of locations (two as an example, at intervals in the longitudinal direction (axial direction, conveyance direction) of the inspection target 100). Location). In these inspection units 10, the image pickup unit 3 and the optical component 4 are arranged differently. In the inspection unit 10U located on the rear side (the left side in FIG. 1) in the transport direction, the imaging unit 3U(3) is located on the upper side in FIG. 1 and the optical component 4 is located on the lower side. Therefore, the imaging unit 3U of the inspection unit 10U acquires an image of the lower region (imaging region A) of the inspection object 100 in FIG. On the other hand, in the inspection unit 10D located on the front side (right side in FIG. 1), the imaging unit 3D(3) is located on the lower side in FIG. 1, and the optical component 4 is located on the upper side. Therefore, the imaging unit 3D of the inspection unit 10D acquires an image of the upper region (imaging region A) in FIG. 1 of the inspection object 100. That is, the plurality of imaging units 3 respectively image different regions (sites, positions) on the surface 100a of the inspection object 100. The transport direction or the direction opposite to the transport direction is an example of the first direction.

  The imaging unit 3 is a line sensor. The imaging unit 3 has a plurality of photoelectric conversion elements (not shown) arranged in a line along the width direction of the inspection object 100. That is, the imaging unit 3 acquires a linear image along the width direction of the inspection object 100. Here, the image may also be referred to as image data, brightness value data for each pixel corresponding to each photoelectric conversion element, or brightness value data string. In each image sensor, for example, the data of the brightness value is obtained with 256 gradations. The image capturing unit 3 acquires a one-dimensional image at each captured time. In the case of the monochrome (black and white) image pickup unit 3, one image data is acquired for each image pickup device, and in the case of the color image pickup unit 3, a plurality of image pickup devices (for example, R (red), G (green), B) are obtained for each image pickup device. (Blue) three) image data is acquired.

  FIG. 3 is an explanatory diagram of the image IA1 and the image IA2 acquired by the inspection device 1. Irradiation of light onto the inspection object 100 by the light sources 2U and 2D is alternately executed. Then, the image pickup by the image pickup unit 3 and the irradiation of the inspection object 100 by the light sources 2U, 2D, that is, the switching of the light emission of the light sources 2U, 2D are synchronized. As illustrated in FIG. 3, the imaging unit 3 includes the one-dimensional image IL1 of the inspection object 100 when illuminated from the obliquely rear (rear side) with respect to the transport direction by the light from the light source 2U and the light source 2D. And the one-dimensional image IL2 of the inspection object 100 when illuminated from the front (front side) obliquely with respect to the transport direction. The image IL1 is an example of a first image, is referred to as a first line image, and is represented by “1” in FIG. The image IL2 is an example of a second image, is referred to as a second line image, and is represented by “2” in FIG. The image processing unit 204 (see FIGS. 6 and 7) can arrange the images IL1 of the inspection object 100 when they are illuminated by the light from the light source 2U in the order of acquisition to obtain the two-dimensional image IA1 and the light source 2D. The two-dimensional image IA2 can be obtained by arranging the images IL2 of the inspection object 100 when illuminated by the light from the acquisition order. The image IA1 is an example of the first image and may also be referred to as a first area image. The image IA2 is an example of a second image and may also be referred to as a second area image.

  The switching frequency of the light from the light sources 2U and 2D, that is, the imaging frequency of each line by the imaging unit 3 is set to a relatively high value (for example, 6 kHz). Therefore, the images IA1 and IA2 can be made similar to the image captured by the area sensor (the image capturing unit that captures a two-dimensional area) with the surface 100a (but half) of the inspection target 100 still. The line sensor has relatively higher resolution than the area sensor and also has relatively high responsiveness. Therefore, by using the line sensor, it may be possible to detect an abnormality (inspection) with higher accuracy.

  FIG. 4 is an exemplary schematic diagram of the image IA1 acquired by the inspection device 1, that is, the image IA1 obtained by the process of FIG. Since the imaging unit 3U acquires an image of the inspection object 100 via the plurality of mirrors 4L and 4R, as illustrated in FIG. 4, it is obtained from the plurality of line images IL1 captured by one imaging unit 3U. The area image IA1 includes a plurality of images Ia and Ib of the inspection object 100. Referring to FIG. 2, these images Ia and Ib deviate from the line L (see FIG. 2) connecting the imaging unit 3U and the inspection target 100 on the opposite side of the inspection target 100 from the side opposite to the imaging unit 3U. It can be seen that the images are equivalent to the images viewed from different positions, and that the images Ia and Ib are images at different positions on the surface 100a of the inspection object 100. The area image IA1 includes an image In that does not pass through the mirrors 4L and 4R of the inspection object 100 between the image Ia and the image Ib. However, the focus of the imaging unit 3U is set corresponding to the images Ia and Ib, so the image In is blurred. Therefore, the image In is not used for abnormality detection. The background of these images Ia, Ib, In is set dark. The image processing unit 204 executes the image processing on the image Ia in the area images IA1 and IA2 to detect an abnormality in the place where the image Ia of the front surface 100a was captured by the inspection unit 10U, and at the same time, the area images IA1 and IA2. By performing the image processing on the inside image Ib, the abnormality is detected in the place where the image Ib of the front surface 100a was captured in the inspection unit 10U. Then, the image processing unit 204 performs the same image processing on the area images IA1 and IA2 obtained from the line image captured by the other image capturing unit 3D, so that the image of the front surface 100a in the inspection unit 10D. Abnormalities are also detected at the place where Ia was taken and the place where the image Ib of the surface 100a was taken in the inspection unit 10D. As described above, the arrangement of the imaging unit 3 and the optical component 4 is reversed between the inspection unit 10U and the inspection unit 10D. Therefore, the image processing unit 204 can detect abnormality at four different locations in the circumferential direction of the surface 100a.

  FIG. 5 is a block diagram of the inspection device 1. As shown in FIG. 5, the inspection apparatus 1 includes, for example, a control unit 20 such as a central processing unit (CPU), a read only memory (ROM 21), a random access memory (RAM 22), a solid state drive (SSD 23), A light irradiation controller 24, an imaging controller 25, a transport controller 26, a display controller 27, etc. are provided. The inspection device 1 may include a hard disk drive (HDD) instead of the SSD 23.

  The light irradiation controller 24 controls light emission and light extinction (light emission stop) of the light sources 2U and 2D based on the control signal from the control unit 20. When a variable device such as a shutter that switches between opening and closing to switch between transmission and blocking of light is provided, the light irradiation controller 24 controls the operation of the variable device. The imaging controller 25 controls the imaging by the imaging unit 3 based on the control signal from the control unit 20. The transport controller 26 controls the transport device 30 based on the control signal received from the control unit 20, and controls the transport of the inspection target 100, for example, the start, stop, speed, etc. of the transport. The display controller 27 controls the display device 40 based on the control signal from the control unit 20. Further, the control unit 20 reads out and executes a program (application) installed in the ROM 21, SSD 23 or the like as a non-volatile storage unit. The RAM 22 temporarily stores various data used when the control unit 20 executes the program and executes various arithmetic processes. Note that the hardware configuration shown in FIG. 5 is merely an example, and various modifications such as a chip or a package can be implemented. Further, various arithmetic processes can be performed in parallel, and the control unit 20 and the like can be configured as hardware capable of performing parallel processing.

  FIG. 6 is a block diagram of the control unit 20, and FIG. 7 is a block diagram of the image processing unit 204. The control unit 20 and the image processing unit 204 function (operate) as at least a part of the inspection device 1 in cooperation with hardware and software (program).

  As shown in FIG. 6, the control unit 20 functions as a light irradiation control unit 201, an imaging control unit 202, a conveyance control unit 203, an image processing unit 204, a display control unit 205, and the like. The light irradiation controller 201 controls the light irradiation controller 24. The imaging control unit 202 controls the imaging controller 25. The transport control unit 203 controls the transport controller 26. The image processing unit 204 performs image processing on the image data acquired by the imaging unit 3. By the image processing by the image processing unit 204, the abnormality of the inspection object 100 is detected. Therefore, the image processing unit 204 is an example of the abnormality detection unit. The display control unit 205 controls the display device 40. The display device 40 is, for example, a liquid crystal display (LCD) or an organic electroluminescent display (OELD).

  As illustrated in FIG. 7, the image processing unit 204 includes an image acquisition unit 204a, a shading correction unit 204b, a difference image acquisition unit 204c, an abnormal candidate region detection unit 204d, an index calculation unit 204e, and an abnormality detection unit 204f. ing.

  Here, a procedure in which the inspection device 1 executes the inspection of the surface 100a of the inspection object 100 having a spiral concave portion or convex portion on the outer periphery will be described. FIG. 8 is a flowchart of a procedure for detecting an abnormality by the inspection device 1. The spiral concave or convex portion is, for example, a groove, a protrusion, a step, or the like, and these are formed when a spiral wound material such as a tape, a ribbon, or a wire is provided. Easy to be affected. The wound material is wound in the longitudinal direction at a constant pitch, for example. The wound material includes one that is exposed on the outer circumference and one that is not exposed. When the wound material is not exposed to the outer periphery and the wound material is provided inside the outer wall (wall portion), the portion that covers the wound material spirally protrudes.

  The image processing unit 204 first functions as the image acquisition unit 204a and acquires the images IA1 and IA2 (S10 in FIG. 8).

  FIG. 9 shows the image Ia in the image IA1, and FIG. 10 shows the image Ia in the image IA2. In the image IA1 of FIG. 9, one of the width directions of the image Ia (right side in FIG. 9) is bright, and in the image IA2 of FIG. 10, the other width direction of the image Ia (left side in FIG. 9) is bright. As described above, in the images IA1 and IA2 of the inspection object 100 having the spiral concave or convex portion on the outer periphery, a change in luminance (unevenness of luminance) due to the location due to the spiral shape occurs, and further, the longitudinal direction. Due to the reversal of the light irradiation direction in, the phenomenon occurs that bright places in the images IA1, IA2 are reversed in the lateral direction.

  As a result of intensive studies by the inventors, the unevenness in the surface 100a has a difference in brightness between the image IA1 and the image IA2. It has been found that there is something that can be detected as. However, in the image IA1 and the image IA2, when the uneven brightness is generated due to the spiral shape of the surface 100a as described above, the difference in brightness due to the spiral shape is generated in the difference image, and thus the difference in brightness is relatively large. It is not possible to discriminate whether the area is caused by the unevenness defect or the spiral shape, and the accuracy of abnormality detection may be reduced.

  Therefore, as a result of intensive studies by the inventors, by performing shading correction on the images IA1 and IA2 respectively, it is possible to reduce the unevenness of brightness associated with the spiral shape in the images IA1 and IA2, and the shading-corrected image IA1′ is obtained. , IA2' (FIGS. 11 and 12) and their difference image IS (FIG. 13), it was found that the effect of the brightness difference due to the spiral shape can be reduced. In the following, the shading-corrected image IA1' is referred to as a shading image IA1', and the shading-corrected image IA2' is referred to as a shading image IA2'. The shading image IA1' is an example of the first shading image, and the shading image IA2' is an example of the second shading image.

  Therefore, the image processing unit 204 functions as the shading correction unit 204b after S10 of FIG. 8 and applies shading correction to the images IA1 and IA2, respectively (S11 of FIG. 8). The shading parameters in S11 can be adjusted appropriately.

  FIG. 11 shows the image Ia in the shaded image IA1', and FIG. 12 shows the image Ia in the shaded image IA2'. By comparing FIG. 11 with FIG. 9, comparing FIG. 12 with FIG. 10, and comparing FIG. 11 with FIG. 12, shading correction is performed on each of the images IA1 and IA2 to obtain the luminance associated with the spiral shape. It can be seen that the unevenness can be reduced.

  Next, after S11 of FIG. 8, the image processing unit 204 functions as a difference image acquisition unit 204c and acquires a difference image IS between the shading image IA1' and the shading image IA2' (S12 of FIG. 8). The difference image between the shading image IA1' and the shading image IA2' is an image of the absolute value of the difference, and hereinafter, this is referred to as the shading difference image IS. The shading difference image IS is an example of a difference image.

  FIG. 13 is an image Ia in the shading difference image IS. From FIG. 13, in the shading difference image IS, there is a region Id1 having a larger absolute value of the difference than the other regions corresponding to the unevenness defect in the region D1 surrounded by the broken line, and the region Id is regarded as the unevenness defect. It turns out that it can be detected.

  In order to detect such an abnormality, the image processing unit 204 functions as the abnormality candidate region detection unit 21d after S12 of FIG. 8 and, for example, binarizes, groups, and labels the shading difference image IS. Is performed, and the absolute difference is equal to or greater than the difference threshold (binarization threshold), the area is composed of a plurality of adjacent pixels, and the area (number of pixels) of the area is the area threshold (pixel number threshold). The labeled area as described above is detected as an abnormality candidate area (S13 in FIG. 8).

  By the way, FIG. 14 shows the image Ib in the shaded image IA1′, FIG. 15 shows the image Ib in the shaded image IA2′, and FIG. 16 shows the shaded image IA1′ in FIG. 14 and the shaded image in FIG. It is the image Ib in the shading difference image IS obtained from IA2′. In each of FIGS. 14 to 16, there is a portion Id2 having a larger absolute value of the difference than the other portion corresponding to the marking (laser marking) in the area D2 surrounded by the broken line, and the portion Id2 is used as the marking. It turns out that it can be detected. However, since the marking is not abnormal, it is convenient to detect the marking separately from the unevenness defect. In addition, dirt is detected in the same manner as marking, but dirt is different in correspondence from the unevenness defect, so it is convenient to detect it separately from the unevenness defect.

  Therefore, as a result of intensive studies, the inventors have found that an abnormality candidate region based on laser marking or stains and an abnormality candidate region based on irregularities can be distinguished by two characteristics.

  FIG. 17 is a plot diagram showing two features for each abnormality candidate region. In FIG. 17, the horizontal axis represents the average value of the brightness values in the shading image IA1′ in the abnormality candidate region detected in S13, and the average value of the brightness values in the shading image IA2′ in the abnormality candidate region detected in S13. Of these, the lower value X. Here, the average value of the brightness values in the shading image IA1′ in the abnormality candidate region is an example of the first representative value, and the average value of the brightness values in the shading image IA2′ in the abnormality candidate region is the second representative value. Is an example. The first representative value and the second representative value are not limited to average values. The value X is an example of the first index.

  In addition, in FIG. 17, the vertical axis represents the average value (representative value) Y of the luminance values (absolute difference value) of each pixel in the shading difference image IS in the abnormality candidate region detected in S13. The average value Y of the brightness values of each pixel in the shading difference image IS is an example of the third representative value and the second index. The third representative value is not limited to the average value. For example, the first representative value, the second representative value, and the third representative value may all be representative values different from the average value such as the median value.

  By analyzing a large number of samples, the inventors have found that the abnormal candidate region as marking or stain (black circle in FIG. 17) is X>Th1 and Y<Th2, and thus the abnormal candidate region as irregularity defect (FIG. 17). It has been found that the white circle in the middle) satisfies X≦Th1 or Y≧Th1. The threshold Th1 is an example of a first threshold, and the threshold Th2 is an example of a second threshold.

  In order to perform such a distinction, the image processing unit 204 functions as the index calculation unit 204e after S13 in FIG. 8 and calculates the above-described value X and average value Y (S14 in FIG. 8). Then, the image processing unit 204 functions as the abnormality detection unit 204f, and based on the comparison between the value X and the threshold value Th1 and the comparison between the average value Y and the threshold value Th2, the image processing unit 204 distinguishes from the abnormality candidate region as marking or stain. Separately, the abnormality candidate area having the unevenness defect is detected as an abnormality (S15 in FIG. 8). The index calculation unit 204e is an example of the first index calculation unit and is also an example of the second index calculation unit.

  It should be noted that in S15, the abnormal candidate region as a marking or stain where X>Th1 and Y<Th2 is distinguished from the abnormal candidate region as an unevenness defect where X≦Th1 or Y≧Th2. Is not limited. For example, it may be distinguished by the size of Y with respect to the discriminant function F(X)=a·X+b shown in FIG. In this case, the abnormality candidate area with F(X)<Y is an abnormality candidate area as marking or stain, and the abnormality candidate area with F(X)≧Y is an abnormality candidate area as unevenness defect. , Can be distinguished. Further, for example, the abnormal candidate area where the value X and the average value Y are within the closed loop Lc of the broken line such as the ellipse in FIG. 17 is the abnormal candidate area as marking or stain, and the value X and the average value Y are the closed loop Lc. The abnormality candidate area that is on the upper side or outside the closed loop Lc may be distinguished as the abnormality candidate area having the unevenness defect. That is, in S15, the abnormality detection unit 204f detects an abnormality candidate region in which the first index and the second index satisfy the predetermined condition as an abnormal region.

  As described above, in the present embodiment, the abnormality detection unit 204f determines from the shading difference image IS (difference image) between the shading image IA1′ (first shading image) and the shading image IA2′ (second shading image). Detect an abnormality. Thereby, for example, it is possible to reduce or eliminate the unevenness of the brightness depending on the place in the image, and thus it is possible to detect the defect (abnormality) with higher accuracy.

  In addition, the inspection object 100 is a long inspection object 100 having a spiral concave or convex portion on the outer periphery, and is a bright region of the image IA1 (first image) and the image IA2 (second image). Is the inspection object 100 in which the dark area and the dark area are reversed in the direction intersecting the longitudinal direction (short direction). As described above, the anomaly detection unit 204f detects an anomaly from the shading difference image IS, so that it is possible to suppress erroneous detection based on the uneven brightness caused by the spiral shapes generated in the images IA1 and IA2.

  Further, the abnormality detection unit 204f detects an abnormality candidate region in which the value X (first index) and the average value Y (second index) satisfy a predetermined condition as an abnormal region. In addition, the abnormality detection unit 204f selects an abnormality candidate area in which the value X is smaller than the threshold Th1 (first threshold) or the average value Y is larger than the threshold Th2 (second threshold) among the detected abnormality candidate areas. , As an abnormal area. Thereby, the irregularities can be detected separately from the marking or the stain.

  Although the embodiments of the present invention have been illustrated above, the above embodiments are merely examples. The embodiment can be implemented in various other forms, and various omissions, replacements, combinations, and changes can be made without departing from the spirit of the invention. In addition, the configuration and shape of the embodiment may be partially replaced with other configurations and shapes. Also, the specifications (structure, type, direction, angle, shape, size, length, width, thickness, height, number, arrangement, position, material, etc.) of each configuration and shape should be changed appropriately. Can be implemented. For example, the inspection object may be something other than the long component. Further, the inspection device can detect an abnormality other than the above-mentioned abnormality by image processing. Further, the inspection device may perform the inspection on the image acquired in a stationary state. Further, the optical component may be other than a mirror (for example, a lens, a prism, etc.). Also, three or more parts of the inspection object may be included in the image. Further, the image capturing section may acquire a linear image extending in a direction intersecting the width direction (oblique direction). Further, three or more image capturing units may be provided. The discriminant function is not limited to the linear function.

  1... Inspection device (image inspection device), 204a... Image acquisition unit, 204b... Shading correction unit, 204d... Abnormal candidate region detection unit, 204e... Index calculation unit (first index calculation unit, second index calculation unit), 204f ... anomaly detection part, 100... inspection object, 100a... front surface (inspection surface), IA1... image (first image), IA2... image (second image), IA1'... shading image (first shading image), IA2'... Shading image (second shading image), IS... Shading difference image (difference image), Th1... Threshold value (first threshold value), Th2... Threshold value (second threshold value).

Claims (4)

A first image photographed in a state in which light is emitted from the rear side of the first direction along the longitudinal direction of the inspection target on the inspection target surface that is the surface of the elongated inspection target, and the inspection target surface An image acquisition unit that acquires a second image captured in a state where light is emitted from the front side in the first direction,
A shading correction unit that acquires a shading-corrected first shading image from the first image and acquires a shading-corrected second shading image from the second image,
An anomaly detection unit that detects an anomaly from a difference image between the first shading image and the second shading image,
An image inspection apparatus equipped with.   The inspection object is a long inspection object having a spiral concave or convex portion on the outer periphery, and a bright region and a dark region in the first image and the second image are in the longitudinal direction. The image inspection device according to claim 1, wherein the image inspection device is an inspection object that is reversed in a direction in which it intersects. An abnormality candidate area detection unit that detects an abnormality candidate area from the difference image,
The lower one of the first representative value of the brightness values in the abnormality candidate region of the first shading image and the second representative value of the brightness values of the abnormality candidate region in the second shading image is set to the first value of the abnormality candidate region. A first index calculation unit that calculates as an index,
The third representative value for the absolute value of the difference value for each pixel between the luminance value in the abnormality candidate region of the first shading image and the luminance value in the abnormality candidate region of the second shading image is set to the third representative value of the abnormality candidate region. A second index calculation unit that calculates as two indexes,
Equipped with
The image inspection apparatus according to claim 1, wherein the abnormality detection unit detects, as an abnormal region, the abnormality candidate region in which the first index and the second index satisfy a predetermined condition.   The image according to claim 3, wherein the abnormality detection unit detects, as the abnormal area, the abnormality candidate area in which the first index is smaller than a first threshold value or the second index is larger than a second threshold value. Inspection equipment.