JP4810659B2 - Appearance inspection apparatus, appearance inspection method, appearance inspection program, and information recording medium recording the same - Google Patents

Description

  The present invention belongs to a technical field related to an appearance inspection apparatus for inspecting a surface state of an object.

  In recent years, products that are incorporated into a finished product together with other parts such as parts have come to be attached to other parts or finished products by bonding, joining, etc. Since even the scratches on the surface affect the installation, these products are required to have high quality without defects such as scratches.

  In particular, recently, when an appearance inspection is performed on an object having a shape that is an object of an axis based on a rotation axis (hereinafter referred to as an “inspection object”), the inspection object is illuminated by a light source from above. Under such conditions, an inspection apparatus is known in which an inspection target is imaged by an imaging camera from the central axis direction, and an appearance inspection is performed based on the captured still screen.

Specifically, the inspection apparatus performs predetermined processing on the captured still screen, and inspects for the presence or absence of defects such as dirt, dents, and scratches on the surface of the inspection target. Based on the inspection result, the quality of the inspection object is determined (for example, Patent Document 1).
Japanese Patent Laid-Open No. 5-72141

  However, in the above-described inspection apparatus, the reference image for determining the inspection is a still image. Therefore, when noise is included in the captured still image, the inspection is accurately performed. I can't.

  In addition, the inspection apparatus binarizes the brightness of each pixel constituting the captured image and detects the presence or absence of defects on the surface based on the binarized brightness of each pixel. Since the lighting conditions for capturing an image are strictly required, that is, the luminance value when binarization changes depending on the lighting conditions, it is weak against environmental changes and is absolute An inspection environment is required.

  The present invention has been made to solve the above-described problems, and its object is to recognize the luminance of each pixel as a relative change based on a plurality of captured images, and is resistant to environmental changes. It is another object of the present invention to provide an appearance inspection system that can determine the quality of an inspection target.

  In order to solve the above-described problem, the invention described in claim 1 is based on an image of an object to be inspected having a shape that is an axis object based on a predetermined axis by an imaging camera device and based on the captured image data. A visual inspection system for performing visual inspection of the inspection object, the rotation device rotating the inspection object relative to the imaging camera device based on the predetermined axis, and the rotation of the rotation means In conjunction with the imaging camera device that images the inspection object from the predetermined axis direction at every predetermined rotation angle, and for each image data output from the imaging camera device, given image data and Based on the image data captured at the previous rotation angle of the given image data, the image in the target region to be inspected of the object to be inspected with reference to the luminance in each pixel constituting the image. An image generation means for generating a difference image indicating a difference image between the detection means, a detection means for detecting the number of pixels having a luminance change based on the generated difference images, and the detection means based on the detected number of pixels. Determination means for determining the presence or absence of defects on the appearance of the inspection object.

  With this configuration, the invention described in claim 1 generates, for each image data, a difference image in the target region on the basis of the luminance in the pixel, and the relative change based on the luminance of each pixel in the difference image. Therefore, when there is an abnormality such as a scratch or dust in the target area, only the luminance of the pixel in the part with the abnormality changes, and the abnormality in the target area is determined by the relative luminance of each pixel. Change can be detected.

  That is, the invention described in claim 1 does not perform inspection only based on one image obtained by imaging the appearance of an object to be inspected, but finds a luminance difference between images by generating a difference image, Since the non-uniform portion of the inspection object is specified, image processing based on relative values can be executed instead of image processing based on absolute values.

  Therefore, since the invention according to claim 1 can determine the presence or absence of defects on the appearance of the inspection object based on the relatively changing luminance, the inspection of the lighting environment or the like during the imaging of each image Even if there is a change in the above environment, it is possible to accurately determine the quality of the inspection object without affecting the inspection result.

  As a result, the invention described in claim 1 is not required to strictly prepare the inspection environment, is easy to install, and allows the inspector to easily check the inspection status, and is resistant to environmental changes, and The installation environment can be made flexible, and the cost for improving the installation environment can be reduced.

  Further, the invention according to claim 2 is a non-target region which is a non-target region for inspection of the inspection object when the image generation unit generates the difference image in the inspection system according to claim 1. The filter processing for making it impossible to detect the luminance component of the image belonging to the image processing unit is performed.

  With this configuration, the invention according to claim 2 can accurately generate the difference image based only on the inspection target and is not affected by the non-target region, so that the quality of the inspection object can be accurately determined. Can be determined.

  The invention according to claim 3 is the inspection system according to claim 1, wherein when the image generation unit generates the difference image, the target region is divided into a plurality of regions, and each region is divided into a plurality of regions. The filter processing for adjusting the luminance level to be detected in the luminance component of the image belonging to the target area is performed.

  With this configuration, the invention described in claim 3 introduces the concept of sensitivity when detecting luminance in the target region, and sets the luminance level detected as a luminance change in each pixel constituting the difference image in each region. Can be set for each.

  Therefore, the invention according to claim 3 increases the sensitivity in a region where it is difficult to detect an abnormality such as a scratch or dust in the target region, that is, it is difficult to detect a change in luminance. Detection error in the target area by adjusting the sensitivity when detecting the brightness for each area, such as reducing the sensitivity, in areas where it is easy to detect abnormalities such as dust Thus, it is possible to accurately detect the luminance change of the pixel in the target area.

  According to a fourth aspect of the present invention, in the appearance inspection system according to any one of the first to third aspects, the determination means includes a total pixel belonging to the target region in all the image data of each rotation angle. A configuration for determining the presence or absence of defects on the appearance of the inspection object based on the ratio between the number of pixels and the number of detected pixels.

  With this configuration, the invention described in claim 4 eliminates errors that occur when the target area is imaged, such as noise generated on the image when generating the difference image, and accurately detects the pixels in the target area. A change in luminance can be detected.

  According to a fifth aspect of the present invention, an inspection object having a shape that is a target of an axis is imaged by an imaging camera device based on a predetermined axis, and an appearance of the inspection object is based on the captured image data. A visual inspection method for inspecting, wherein a rotation control step of rotating a rotating device relative to the imaging camera device based on the predetermined axis, and a rotation control step in conjunction with rotation in the rotating means. An imaging control step of imaging the object to be inspected by the imaging camera device from the predetermined axis direction for each predetermined rotation angle, for each image data output from the imaging camera device, the given image data and the Based on the image data captured at the previous rotation angle of the given image data, the brightness of each pixel constituting the image is used as a reference to the target area to be inspected of the inspection object. An image generation step of generating a difference image indicating an image of a difference between images, a detection step of detecting the number of pixels having a luminance change based on each of the generated difference images, and based on the detected number of pixels And a determination step for determining the presence or absence of defects on the appearance of the inspection object.

  With this configuration, the invention according to claim 5 generates, for each image data, a difference image in the target area on the basis of the luminance in the pixel, and the relative change based on the luminance of each pixel in the difference image. Therefore, when there is an abnormality such as a scratch or dust in the target area, only the luminance of the pixel in the part with the abnormality changes, and the abnormality in the target area is determined by the relative luminance of each pixel. Change can be detected.

  That is, the invention according to claim 5 does not perform inspection only based on one image obtained by imaging the appearance of the object to be inspected, but finds a luminance difference between the images by generating a difference image. Since the non-uniform portion of the inspection object is specified, image processing based on relative values can be executed instead of image processing based on absolute values.

  Therefore, the invention according to claim 5 can determine the presence or absence of defects on the appearance of the inspection object based on the luminance that changes relatively. Even if there is a change in the above environment, it is possible to accurately determine the quality of the inspection object without affecting the inspection result.

  As a result, the invention according to claim 5 is not required to prepare the inspection environment strictly, is easy to install, and allows the inspector to easily check the inspection status, and is resistant to environmental changes, and The installation environment can be made flexible, and the cost for improving the installation environment can be reduced.

  The invention according to claim 6 or 7 is a computer in which an object to be inspected having a shape that is an axis object based on a predetermined axis is imaged by an imaging camera device, and the image data is captured based on the captured image data. An appearance inspection program for inspecting an appearance of an object to be inspected, wherein the computer rotates a rotation device relative to the imaging camera device with respect to the object to be inspected based on the predetermined axis, Imaging control means for imaging the inspection object by the imaging camera device from the predetermined axis direction at each predetermined rotation angle in conjunction with rotation in the rotation means, and for each image data output from the imaging camera device Based on given image data and image data taken at the previous rotation angle of the given image data, luminance at each pixel constituting the image As a reference, an image generating means for generating a difference image indicating an image of a difference between images in a target region to be inspected of the inspection object, and detecting the number of pixels having a luminance change based on each of the generated difference images And detecting means for determining whether or not there is a defect on the appearance of the inspection object based on the detected number of pixels.

  With this configuration, the invention according to claim 6 or 7 generates, for each image data, a difference image in the target region on the basis of the luminance in the pixel, and relative to the difference image based on the luminance of each pixel. Therefore, if there is an abnormality such as a scratch or dust in the target area, only the luminance of the pixel in the part with the abnormality changes, and the abnormality of the target area is determined by the luminance of each pixel. It can be detected as a relative change.

  That is, the invention according to claim 6 or 7 finds a luminance difference between images by generating a difference image, rather than performing an inspection only based on one image obtained by imaging the appearance of the inspection object. Since the non-uniform portion of the inspection object is specified, image processing based on relative values can be executed instead of image processing based on absolute values.

  Therefore, the invention according to claim 6 or 7 can determine the presence / absence of a defect on the appearance of the inspection object based on the luminance that changes relatively. Even if there is a change in the environment on the inspection, the quality of the inspection object can be accurately determined without affecting the inspection result.

  As a result, the invention described in claim 6 or 7 does not need to prepare the inspection environment strictly, is easy to install, and allows the inspector to easily check the inspection status, so it is resistant to environmental changes. In addition, the installation environment can be flexible, and the cost for maintaining the installation environment can be reduced.

  Since the present invention can determine the presence or absence of defects on the appearance of an object to be inspected based on the luminance that changes relatively, there is a change in the inspection environment such as the illumination environment while each image is captured. Even so, it is possible to accurately determine the quality of the inspection target without affecting the inspection result.

  Therefore, the present invention eliminates the need for strict preparation of the inspection environment, is easy to install, and allows the inspector to easily check the inspection status, and is resistant to environmental changes and flexible to the installation environment. The cost for maintaining the installation environment can be reduced.

  Next, preferred embodiments of the present invention will be described with reference to the drawings. In addition, this invention includes the range which has the technical feature, and is not limited to drawing shown below.

  In the embodiment described below, the present invention is applied to an appearance inspection system and an appearance inspection method for inspecting appearance of components (injection cups) used for automobile injection (fuel injection device) as an inspection object. It is an embodiment.

  First, the configuration of the appearance inspection system of the present embodiment and the schematic operation of each device constituting the appearance inspection system will be described with reference to FIG. FIG. 1 is a block diagram showing the configuration of the appearance inspection system of the present embodiment.

  The appearance inspection system S of the present embodiment is installed in, for example, a manufacturing process line of an injection cup C that is an inspection object, and determines whether there is a defect on the appearance of the manufactured injection cup C. Yes.

  The injection cup C is incorporated in the injection and used as a part of the device. When the joint cup is joined to another part and joined, if there is a fine flaw on the joint surface, the joint part Will break due to aging, etc., causing significant damage to the injection itself. Therefore, the appearance inspection system S of the present embodiment inspects fine flaws on the joint surface of the injection, and determines the quality.

  As shown in FIG. 1, the appearance inspection system S includes a jig J for installing an injection cup C having a cylindrical shape with an upper opening, and a rotating device 100 that rotates the jig J during inspection. The imaging camera device 200 that captures the appearance of the injection cup C fixed to the jig J, and the appearance of the injection cup C is inspected based on the image data captured by the imaging camera device 200, and whether there is a defect on the appearance. And an inspection apparatus 300 for determining whether or not.

  The rotation device 100 fixes the injection cup C, and uses the cylindrical axis of the injection cup C (the central axis in the height direction of the cylinder) as a reference, and in conjunction with the imaging camera device 200 under the control of the inspection device 300 The cup C is intermittently rotated every predetermined rotation angle.

  In particular, the rotation device 100 of the present embodiment is used to detect the rotation angle of the jig 110 and the motor 110 for rotating the jig J to which the injection cup C is fixed based on the control of the inspection device 300. And an encoder 120 for outputting a pulse to the inspection apparatus 300 at every predetermined rotation angle in the turntable.

  In the present embodiment, it is avoided that an image cannot be clearly captured due to “blurring” due to the moving speed during rotation, and the appearance of the injection cup C is imaged in a state where the rotation of the injection cup C is stopped. It has become. For this reason, the rotation apparatus 100 of this embodiment rotates the jig | tool J intermittently. However, when the imaging camera device 200 having a sufficiently high shutter speed is used, it is not necessary to stop the rotation.

  Further, for example, the rotation device 100 of the present embodiment is configured to intermittently rotate every 12 degrees when 30 still images are taken at one rotation. However, the rotation angle can be varied based on the still image to be captured, and can be arbitrarily set.

  The imaging camera device 200 is arranged to face the upper surface of the injection cup C, and images the appearance of the injection cup C from above the injection cup C. In addition, the imaging camera device 200 captures a plurality of still images for one rotation for each predetermined rotation angle in conjunction with the rotation device 100 under the control of the inspection device 300, and each captured still image is captured. The image data is output to the inspection apparatus 300.

  For each piece of image data output from the imaging camera device 200, the inspection apparatus 300 is based on given image data and image data taken at the previous rotation angle of the given image data. An image (hereinafter, referred to as “difference image”) indicating an image of a difference between images in a target region (hereinafter, simply referred to as “target region”) to be inspected by the injection cup C is generated. . Then, the inspection apparatus 300 calculates the total sum of the generated difference images, calculates the number of pixels in which luminance change has occurred in the totaled image (hereinafter referred to as “total image”), and detects the detected pixel. Based on the number of pixels, the presence / absence of a defect on the appearance of the ignition coil to be inspected is determined.

  In particular, the inspection apparatus 300 according to the present embodiment generates a difference image on the basis of the luminance in each pixel constituting the captured still image, and generates a sum image based on each difference image, thereby changing the pixel change. An image having a trajectory is generated, and when there is an abnormality such as a scratch or dust in the inspection target area, only the luminance in the pixel of the part having the abnormality in the summed image changes. This abnormality can be detected as a relative change in the luminance of each pixel.

  Further, when generating the difference image, the inspection apparatus 300 divides the target area into a plurality of areas, and adjusts the luminance level to be detected in the luminance component of the image belonging to the target area for each area. The filter processing is performed to disable the detection of the luminance component of the image belonging to the region that is not subject to the inspection of the injection cup C (hereinafter simply referred to as “non-target region”).

  As described above, the appearance inspection system S according to the present embodiment is configured such that an inspection target to be inspected, that is, a defect on the appearance of the injection cup C is relative to each pixel in an image in which the appearance of the injection cup CC is captured. Since the determination can be made on the basis of a typical change in luminance, it is resistant to environmental changes, and the quality of the inspection object can be accurately determined.

  Next, the shape of the injection cup C serving as an inspection object in the appearance inspection system S of the present embodiment will be described with reference to FIG.

  As shown in FIG. 2, the injection cup C that is an object to be inspected of this embodiment has a cylindrical shape with an upper opening, and a flange portion 10 for joining to other parts on the upper surface of the cylinder. Further, a tapered portion 30 having a shape in which the diameter gradually decreases conically from the flange portion 10 to the body portion 20 is provided. In particular, the flange portion 10 is provided with a rotation preventing portion 40 for stopping the injection cup C itself during joining.

  Further, a hole 50 having a shape protruding outward from the bottom surface with respect to the cylindrical axis of the injection cup C is provided on the bottom surface of the injection cup C.

  Thus, the injection cup C of the present embodiment has a cylindrical shape, that is, an axial target shape with respect to the center axis of the circle, and can be inspected by the appearance inspection system S of the present embodiment. It can be done.

  In addition, the injection cup C of this embodiment is shape | molded, for example by press work or a shaving process.

  In the present embodiment, the injection cup C is used as an object to be inspected. However, if the injection cup C has a shape that is an axis object based on a predetermined axis, the entire shape is not particularly an axis object. Even in this case, if the region to be inspected has a shape that is an axis target, the region can be inspected.

  Next, the inspection apparatus 300 in the appearance inspection system S of the present embodiment will be described using FIG. 3 and FIG. 4 together with FIG. FIG. 3 is a diagram for explaining a reference image that is a basis for generating a difference image, and is a diagram for explaining a filter image generated based on a partial filter image.

  The inspection apparatus 300 according to the present embodiment includes an input processing unit 305 that performs a predetermined input process in which image data of each still image output from the imaging camera apparatus 200 is input, and outputs the input image data to the monitor 400. An output processing unit 310 for performing predetermined output processing for data, a data storage unit 315 for storing various data used for inspection, and various types of image processing for generating a difference image for each input image data On the appearance of the injection cup C based on the number of pixels of the abnormal pixel, the pixel detection unit 325 that detects a pixel with a scratch or the like as an abnormal pixel based on each generated difference image And a determination processing unit 330 that determines whether the product is acceptable.

  The inspection apparatus 300 includes a motor control unit 335 that controls the motor 110 of the rotating device 100, a pulse counter 340 that detects a pulse output from the rotating device 100 and recognizes the rotation angle of the rotating device 100, and an operation. An operation unit 345 used when performing an input by a person, and a system control unit 350 that controls each unit of the rotation device 100, the imaging camera device 200, and the inspection device 300.

  For example, the image processing unit 320 of the present embodiment constitutes an image generation unit of the present invention, and the calculation unit constitutes a detection unit of the present invention. Further, for example, the determination processing unit 330 of the present embodiment constitutes a determination unit of the present invention.

  The input processing unit 305 receives image data of still images captured at predetermined rotation angles, performs predetermined processing on the input image data, and is set in advance. It is designed to convert to a data format.

  For example, when the input image data is analog data, the input processing unit 305 of the present embodiment performs predetermined processing such as processing to convert to digital data or adjustment of image quality in a captured image. .

  Each output processed image data is input to the output processing unit 310, and the output processing unit 310 performs predetermined output processing such as converting each input image data into analog data. The output image data is output to the monitor 400.

  Various data used in the inspection apparatus 300 are stored in the data storage unit 315. For example, various parameters for generating a filter for performing filter processing, image data used as a filter, data of each difference image, Data on the total image or data on the inspection result is stored.

  The data storage unit 315 is configured by, for example, an HDD (Hard Disc Drive) and can be connected to various communication terminal devices (not shown) such as a personal computer based on a communication line such as a network. Yes.

  The image processing unit 320 uses a reference image generation unit 321 that generates a base image (hereinafter referred to as “reference image”) when generating a difference image for each input image data, and a reference image using a filter. And a filter processing unit 322 for generating a difference image.

  For each input image data, the reference image generation unit 321 configures an image based on the input image data and image data captured at a rotation angle immediately before the image data. An image serving as a base is generated when a difference image is generated on the basis of the luminance in the pixel.

Specifically, the reference image generation unit 321 according to the present embodiment assumes that images composed of time-sequentially input image data are S _{1 to} S _f (if the number of shots is 30, f = 30 a), and by the input image data before one step with respect to the image formed S _i and the time-series manner the given image data at a given image data image S ₍ Based on the two images of _i-1) , the difference value of the luminance is calculated for each pixel using the following (formula 1), and the reference image Δ _i is generated.

Note that “i” indicates the reference image Δ calculated based on the i-th input image, and “n” indicates the total number of pixels on the image. Further, “δ _i ” is a difference value in each pixel of the reference image Δ _i and satisfies the following (Expression 2). Furthermore, the image S _i and the image S _(i−1) satisfy the following (Expression 3) and (Expression 4). However, “j” indicates an arbitrary pixel position (1 to n) on the image. This notation is a case where coordinates are expressed in a raster format, and can be expressed as x, y.

  For example, image data D1 having a scratch on the surface of the injection cup C shown in FIG. 3A is input, and image data captured at the rotation angle immediately before the input image data D1 is shown in FIG. ), The reference image generation unit 321 of this embodiment generates a difference image based on the luminance of each pixel constituting the image based on the image data D1 and D2. A reference image B shown in FIG. 3C, which is a basis for this, is generated.

  The filter processing unit 322 generates an image (hereinafter referred to as “filter image”) used for the filter processing based on the parameters stored in advance in the data storage unit 315, and uses each of the generated filter images to generate each reference image. Is subjected to filter processing to generate a difference image.

  Specifically, the filter processing unit 322 of the present embodiment makes it impossible to detect the luminance of each pixel in the non-target region of the difference image based on the parameters stored in advance in the data storage unit 315, and the target A filter image for adjusting the luminance level of each pixel to be detected in the region is generated.

For example, as shown in FIG. 4, the filter processing unit 322 according to the present embodiment includes a partial filter image m ₁ having sensitivity (T ₁ = 0.38) in the target area A and a target area B based on the parameters. a partial filtered image _{m 2} having sensitivity _(T 2 = 0.95), to generate a partial filtered image _{m 3} having a sensitivity _(T 3 = 0.28) in the target area C, respectively, were synthesized by a logical aND operation The filter image M is generated.

  The sensitivity T indicates a ratio when the maximum luminance value in the pixel is “1”. As will be described later, the difference is detected as a difference value even if the ratio is small. In the present embodiment, the detection sensitivity increases as the ratio decreases.

  Further, the filter processing unit 322 of the present embodiment may generate a filter image based on one partial filter image, or may generate a filter image based on four or more partial filter images. .

  On the other hand, when the filter image M is generated, the filter processing unit 322 according to the present embodiment removes non-target regions other than the target region in each reference image generated by the filter image M as described above, For the pixels in the region, each difference image is generated by performing a filtering process using the corresponding sensitivity set as described above as a threshold value.

For example, the filter processing unit 322, as shown in (Equation 5), on the basis of the filter image M, the pixel is set to a non-target region in the calculated reference image delta _i as described above, "m _{In addition} to multiplying “ ₀ = 0”, the pixels set in the target region in the reference image Δ _i are subjected to the first filter processing of multiplying “m ₁ = 1”.

“N” indicates the total number of pixels on the image. “Β _i ” is a reference image (hereinafter referred to as “intermediate image”) calculated by performing the first filter process in the process of calculating the difference image, and “b _{(i, j)} ”. "Indicates the luminance of each pixel of the intermediate image β _i . Furthermore, the luminance b _{(i (} pixel of the _{non-target area)) of the pixel in the non-target area} and the luminance b _{(i, (target area))} of the pixel of the target area in the intermediate image β _i are (Expression 6) and (Expression 6). 7) is satisfied. However, “j” indicates an arbitrary pixel position (1 to n) on the image.

The filter processing unit 322 is set for each pixel of the intermediate image β _i calculated in the filter processing and each pixel of the target area in the filter image M for each pixel of the intermediate image β _i . The sensitivity T is compared with each other, and as shown in (Equation 8) and (Equation 9), if each pixel is larger than the sensitivity T, the luminance of each pixel is set to “1”, and each pixel is less than the sensitivity T. In the case of (2), the second filter processing for setting the luminance of the pixel to “0” is performed, and the difference image _Di is calculated.

Note that “i” indicates the difference image D calculated based on the i-th input image, and “n” indicates the total number of pixels on the image. Further, “d _{(i, j)} ” indicates the luminance of each pixel of the difference image D _i . However, “j” indicates an arbitrary pixel position (1 to n) on the image.

  As described above, the image processing unit 320 of the present embodiment generates the difference image by calculating the difference image while performing the filtering process via the reference image and the intermediate image.

  Note that the image processing unit 320 of the present embodiment performs the filtering process, for example, when a change in luminance is detected for a pixel group in which one pixel or several pixels are adjacent, the pixel or the pixel group. Is determined as noise, and is erased, that is, is set to “0”, similarly to pixels belonging to the non-target region.

  In addition, the image processing unit 320 of the present embodiment may generate a difference image from each piece of image data input directly without performing the above-described filter processing. That is, in this case, the image processing unit 320 generates the above-described reference image as a difference image.

  Each difference image finally calculated by the filter processing unit 322 is input to the pixel detection unit 325, and the image detection unit detects scratches or the like based on each input difference image. An existing pixel is detected as an abnormal pixel.

  Specifically, the pixel detection unit 325 according to the present embodiment uses the (Equation 10) to calculate the sum image A that is the sum of the difference images.

In (Expression 10), “n” indicates the total number of pixels on the image, and “a _i ” is a difference value in each pixel of the summed image A. Further, this total image A can be considered as a histogram by treating j as the location of the flaw and its frequency as the number of flaws, and in this case, the highest value is (f-1). In other words, when scratches are made, scratches are often distributed over a certain range rather than just a single point hitting with a needle, and in the area where several scratches are distributed, the same trajectory is passed. Therefore, in this embodiment, when the scratches pass through the same locus (location), the number of scratches that have passed through the same place is counted rather than simply treated as “1” based on the presence or absence of the locus. By treating as “2”, the information of flaw detection can be strengthened. Furthermore, in the present embodiment, the detection target is rotated once, and the above-described processing is performed. However, in order to improve the accuracy of scratch detection, if the detection target is rotated several times, the scratches that have passed through the same place can be further increased. Since counting can be performed, the accuracy of flaw detection can be improved.

  Then, as shown in (Equation 11), the pixel detection unit 325 detects pixels indicating the value “1” in the total image A as abnormal pixels, and calculates the total number X of the detected number of pixels. It is like that.

  The determination processing unit 330 calculates the sum of the number of pixels in the target area in each difference image, that is, the total number of pixels (hereinafter referred to as “total number of pixels”) obtained by summing the number of pixels in each generated difference image, The presence or absence of a defect on the appearance of the inspection object is determined based on the ratio of the number of pixels detected by the detection unit 325.

  Specifically, the determination processing unit 330 of the present embodiment adds the number of pixels belonging to the target region in each difference image to calculate the total number of pixels, and calculates the total number of pixels and abnormal pixels by (Equation 12). A ratio e of the number of detected pixels is calculated.

  “X” represents the total number of pixels calculated by the pixel detection unit 325, and “R” represents the total number of pixels.

  Further, the determination processing unit 330 compares the calculated ratio e with a threshold value stored in advance in the data storage unit 315 (hereinafter referred to as “determination threshold value”), and the injection cup to be inspected. The presence / absence of a defect on the appearance of C, that is, determination of pass, fail, and re-inspection is performed.

  In particular, when the detection ratio e is smaller than a determination threshold (hereinafter referred to as “pass threshold”) used for passing an inspection, the determination unit 330 determines that the appearance inspection has passed and fails the inspection. When the detection ratio e is larger than a determination threshold value (hereinafter, referred to as “failure threshold value”) used as an object, it is determined that the appearance inspection has failed. In addition, the determination processing unit 330 determines that re-examination is performed when it is equal to or higher than the pass threshold and equal to or lower than the fail threshold.

  For example, in the present embodiment, based on the determination, an injection cup C to be inspected is classified in association with the determination type by an apparatus (not shown) or the like.

  The motor control unit 335 controls the rotation of the rotation device 100 and its stop while interlocking with the imaging camera device 200 when performing an appearance inspection under the control of the system control unit 350.

  The pulse counter 340 receives a pulse output from the encoder 120 when the jig J rotates at a certain angle from the rotating device 100 when performing an appearance inspection under the control of the system control unit 350, and receives the received pulse. The rotation angle is recognized by counting the number of pulses.

  The operation unit 345 is configured by an input interface such as various confirmation buttons, operation buttons for inputting each operation command, a keyboard having a numeric keypad and other numeric keys, or a mouse as an input pointing device, and performs various settings or operation instructions. It is used when performing.

  The system control unit 350 is mainly configured by a central processing unit (CPU) and a ROM / RAM, and comprehensively controls general functions of the appearance inspection system S. The system control unit 350 reads a control program stored in the ROM, and temporarily stores data being processed in the RAM, and performs various processes based on the read control program. Is supposed to run.

  In particular, in the present embodiment, the system control unit 350 controls the start and end of the appearance inspection when performing the appearance inspection of the injection cup C, and generates a filter image, a difference image, and an appearance defect. The process relating to the presence or absence of the process (hereinafter referred to as “inspection process”) is controlled.

  In addition, the system control unit 350 of the present embodiment is connected to each part of the inspection apparatus 300 by a bus B, and is connected to the rotation apparatus 100 or the imaging camera apparatus 200 by a predetermined communication line.

  Next, operation | movement of the test | inspection process in the external appearance test | inspection system S of this embodiment is demonstrated using FIG. FIG. 5 is a flowchart showing the operation of the inspection process in the appearance inspection system S of the present embodiment.

  In the following operation, it is assumed that the injection cup C is already fixedly installed in the rotating device 100, and after the inspection is finished, the injection C is classified by a device (not shown) based on the determination type. Further, this operation is executed for each injection cup C that is an object to be inspected.

  First, the system control unit 350 reads out various parameters stored in the data storage unit 315 such as the number of captured images (number of captured images) and a threshold for determination, and stores them in the internal RAM (step S11). Specifically, the system control unit 350 calculates various values calculated based on various parameters read out or parameters read out while calculating values used for appearance inspection such as a rotation angle when the imaging camera performs imaging. Is set in the RAM.

  Next, the system control unit 350 reads out a parameter for generating a filter image from the data storage unit 315, generates a filter image, and sets it in the filter processing unit 322 (step S12). At this time, the system control unit 350 generates each partial filter image while recognizing the number of images of the partial filter images to be generated, and generates a filter image by combining the generated partial filter images.

  Next, the system control unit 350 controls the rotation device 100 via the motor control unit 335 to start the rotation of the jig J (step S13), and recognizes the rotation angle of the rotation device 100 by the pulse counter 340. It is detected whether or not the rotation device 100 has reached a predetermined rotation angle (step S14).

  Next, when the system control unit 350 detects a predetermined rotation angle in the process of step S14, the system control unit 350 controls the rotation device 100 via the motor control unit 335 and stops the rotation of the jig J (step S15). .

  Next, the system control unit 350 controls the imaging camera device 200 to capture still images in the appearance of the injection cup C (step S16), and determines whether all the still images have been captured, that is, whether capture has been completed. (Step S17).

  At this time, if the system control unit 350 determines that all of the still images have not been captured, the system control unit 350 proceeds to the process of step S13 to start capturing the next still image, and captures all the still images. If it is determined that the capture has been completed, the process proceeds to step S18.

  Next, when it is determined in step S17 that all the still images have been captured, the image processing unit 320 performs the filtering process based on each captured still image as described above. Then, a reference image and an intermediate image are generated to generate each difference image (step S18).

  Next, the system control unit 350 causes the image processing unit 320 to remove noise from each difference image (step S19), and causes the pixel detection unit 325 to calculate a sum image indicating the sum of the difference images (step S20). .

Next, the system control unit 350 causes the pixel detection unit 325 to calculate the total number of pixels detected as abnormal pixels based on the calculated total image (step S21), and causes the determination processing unit 330 to perform the abnormality processing. A ratio between the total number of pixels detected as pixels and the total number of pixels in the target area in each difference image is calculated (step S22).
Next, the system control unit 350 causes the determination processing unit 330 to determine whether or not the calculated ratio is smaller than a pass threshold and whether or not the calculated ratio is greater than a fail threshold (step S23 and Step S24).

  Next, the system control unit 350 causes the determination processing unit 330 to determine that the calculated ratio is smaller than the pass threshold value (step S25), ends the operation, and calculates the calculated value. If it is determined that the ratio is greater than the failure threshold, the failure is determined (step S26), and the operation is terminated. If the determination processing unit 330 determines that the determination processing unit 330 is equal to or greater than the pass threshold and is equal to or less than the fail threshold, the system control unit 350 determines that re-examination is performed, and ends the operation.

  As described above, the appearance inspection system S according to the present embodiment generates a difference image in the target region for each image data on the basis of the luminance of the pixel, and displays a relative change based on the luminance of each pixel in the difference image. Therefore, if there is an abnormality such as a scratch or dust in the target area, only the luminance of the pixel in the part with the abnormality will change, and the abnormality of the target area will be changed relative to the luminance of each pixel. Can be detected as

  That is, the appearance inspection system S of the present embodiment does not inspect only based on one image obtained by imaging the appearance of an inspection object such as the injection cup C, but generates a difference image to generate a difference image. Since the luminance difference is found and the non-uniform portion of the inspection object is specified, the image processing based on the relative value can be executed instead of the image processing based on the absolute value.

  Therefore, the appearance inspection system S of the present embodiment can determine the presence or absence of defects on the appearance of the inspection object based on the relatively changing luminance, so that the lighting environment or the like can be taken while each image is captured. Even if there is a change in the environment on the inspection, it is possible to accurately determine the quality of the inspection object without affecting the inspection result.

  As a result, the appearance inspection system S of the present embodiment does not require a strict inspection environment, is easy to install, and allows the inspector to easily check the inspection status, so it is resistant to environmental changes, In addition, the installation environment can be flexible, and the cost for improving the installation environment can be reduced.

  Further, the appearance inspection system S of the present embodiment can generate a difference image accurately based only on the inspection target by performing the filter process, and also has a concept of sensitivity when detecting the luminance in the target region. Thus, the luminance level detected as a luminance change can be set for each region in each pixel constituting the difference image.

  Therefore, the appearance inspection system S of the present embodiment removes noise in the non-target region, increases sensitivity in a region where it is difficult to detect a change in luminance, and in a region where it is easy to detect a change in luminance. Can adjust the sensitivity when detecting the brightness for each area, for example, by reducing the sensitivity, so that it is possible to accurately detect the change in brightness of each pixel while preventing detection errors in the target area. In addition, the quality of the inspection object can be determined.

  In addition, since the appearance inspection system S of the present embodiment performs image processing by such filter processing, each image processing is extremely simple, and at a general video signal level (30 Hz frame rate). Also in the imaging camera device to be imaged, special hardware-like computation can be eliminated and the image processing is simple and high speed, so that the high-speed imaging camera device or the high-resolution imaging camera device can be realized. Can be easily converted.

  Note that the appearance inspection system S of the present embodiment performs each operation based on the control of the system control unit 350, but a program that defines the operation of the inspection process for performing the appearance inspection is recorded. The above-described recording medium and a computer that reads the recording medium may be provided, and the inspection processing operation similar to the above may be performed by reading the program with this computer.

It is a block diagram showing the composition of one embodiment of the inspection system concerning this application. It is an external view of the hermit table cup of the to-be-inspected object which is the object of inspection by the inspection system in one embodiment. In one embodiment, it is a figure for demonstrating the reference | standard image used as the basis for producing | generating a difference image. In one embodiment, it is a figure for demonstrating the filter image produced | generated based on a partial filter image. It is a flowchart which shows the operation | movement of the test | inspection process in the external appearance inspection system of one Embodiment.

Explanation of symbols

C ... Injection cup (inspection object)
J ... Jig S ... Appearance inspection device 100 ... Rotation device 110 ... Motor 120 ... Encoder 200 ... Imaging camera device 300 ... Inspection device 305 ... Input processing unit 310 ... Output processing unit 315 ... Data storage unit 320 ... Image processing unit 321 ... Reference image generation unit 322 ... Filter processing unit 325 ... Pixel detection unit 330 ... Determination processing unit 335 ... Motor control unit 340 ... Pulse counter 345 ... Operation unit 350 ... System control unit

Claims (7)

An appearance inspection system for imaging an inspection object having a shape to be an axis object based on a predetermined axis by an imaging camera device, and performing an appearance inspection of the inspection object based on the imaged image data,
A rotating device for rotating the object to be inspected relative to the imaging camera device based on the predetermined axis;
The imaging camera device that images the inspection object from the predetermined axial direction at every predetermined rotation angle in conjunction with the rotation of the rotation means;
For each image data output from the imaging camera device, each of the images constituting the image based on the given image data and the image data taken at the previous rotation angle of the given image data Image generation means for generating a difference image indicating an image of a difference between images in a target region to be inspected of the inspection object, based on luminance in the pixel;
A sum image calculating means for calculating a sum image by summing the luminance of each pixel in the generated difference image for each pixel;
Detecting means for detecting the number of pixels having a luminance change based on the total image ;
An appearance inspection system comprising: determination means for determining the presence or absence of a defect on the appearance of the inspection object based on the detected number of pixels. The appearance inspection system according to claim 1,
When the image generation unit generates the difference image, it performs a filter process for disabling detection of luminance components of an image belonging to a non-target region that is a non-target of the inspection of the inspection object. Appearance inspection system. The appearance inspection system according to claim 1,
When the image generation unit generates the difference image, the target region is divided into a plurality of regions, and for each region, a luminance level to be detected in a luminance component of an image belonging to the target region is adjusted. An appearance inspection system characterized by performing filter processing. The appearance inspection system according to any one of claims 1 to 3,
The determination means determines the presence / absence of a defect on the appearance of the inspection object based on a ratio between the total number of pixels belonging to the target area and the number of detected pixels in all image data at each rotation angle. Appearance inspection system. An appearance inspection method for imaging an inspection object having a shape to be an axis object based on a predetermined axis by an imaging camera device, and performing an appearance inspection of the inspection object based on the imaged image data,
A rotation control step of rotating a rotating device relative to the imaging camera device based on the predetermined axis;
An imaging control step of causing the imaging camera device to capture an image of the inspection object from the predetermined axis direction at every predetermined rotation angle in conjunction with the rotation of the rotating means.
For each image data output from the imaging camera device, each of the images constituting the image based on the given image data and the image data taken at the previous rotation angle of the given image data An image generation step of generating a difference image indicating an image of a difference between images in a target region to be inspected of the inspection object, based on luminance in the pixel;
A sum image calculation step of calculating a sum image by summing the luminance of each pixel in the generated difference image for each pixel;
A detection step of detecting the number of pixels having a luminance change based on the total image ;
And a determination step of determining the presence or absence of defects on the appearance of the inspection object based on the detected number of pixels. An appearance inspection program for imaging an inspection object having a shape to be an axis object based on a predetermined axis by a computer using an imaging camera device and performing an appearance inspection of the inspection object based on the captured image data. And
The computer,
A rotation control means for rotating a rotating device relative to the imaging camera device based on the predetermined axis;
An imaging control unit that causes the imaging camera device to capture an image of the inspection object from the predetermined axis direction at every predetermined rotation angle in conjunction with the rotation of the rotation unit.
For each image data output from the imaging camera device, each of the images constituting the image based on the given image data and the image data taken at the previous rotation angle of the given image data An image generating means for generating a difference image indicating an image of a difference between images in a target region to be inspected of the inspection object on the basis of luminance in the pixel;
A sum image calculation means for calculating a sum image by summing the luminance of each pixel in the generated difference image for each pixel;
Detecting means for detecting the number of pixels having a luminance change based on the total image ;
Determination means for determining the presence or absence of defects on the appearance of the inspection object based on the number of detected pixels,
Appearance inspection program characterized by functioning as An information storage medium readable by a computer, wherein the program according to claim 6 is stored.

Images