JP2010175281A - Apparatus for forming image of appearance surface - Google Patents

Abstract

An image representing a state of an appearance surface of an inspection object can be generated with a simple configuration.
An inspection object, which is a rotationally symmetric metal member, is rotated by a drive unit. The line illumination 14 irradiates linear slit light, and the reflected light is imaged by the area camera 18. A predetermined region around the portion representing the reflected light of the linear slit light is cut out from each of a plurality of picked-up images picked up by the image processing unit 32 for each predetermined rotation amount of the inspection object, and the predetermined cut out The image data of the areas are arranged side by side and combined to generate an appearance surface image representing the state of the appearance surface of the inspection object. The defect detection unit 34 detects a part having a luminance value equal to or higher than a predetermined value from the appearance surface image as a defect candidate.
[Selection] Figure 1

Description

  The present invention relates to an appearance surface image generation device, and more particularly to an appearance surface image generation device that generates an image representing the state of the appearance surface of an inspection object.

  Conventionally, a method for measuring a three-dimensional curved surface shape that enables highly reliable measurement in a short time is known (Patent Document 1). In this measurement method, a plurality of linear slit lights are simultaneously rotated and scanned over the entire surface of the object to be measured, the rotational scanning angle of the slit light is measured, and the video signal obtained by imaging the surface of the object to be measured is displayed on the screen. For each position of the surface to be measured corresponding to each pixel, the rotational scanning angle at the moment when one or more slit lights pass through that point or one of the corresponding values is the pixel value. Is composited. Further, the three-dimensional curved surface shape of the measurement target is measured based on the composite image.

JP-A-7-260443

  However, the technique described in Patent Document 1 requires a configuration for rotationally scanning the slit light and a configuration for detecting the rotational scanning angle, which causes a problem that the configuration of the apparatus becomes complicated.

  The present invention has been made to solve the above-described problems, and provides an appearance surface image generation apparatus capable of generating an image representing the state of the appearance surface of an inspection object with a simple configuration. With the goal.

  In order to achieve the above object, an external appearance image generating apparatus according to the present invention includes a light irradiation means for irradiating an external appearance surface of a test object with linear light, and the linear shape from the external appearance surface of the inspection object. Imaging means for imaging a region including transmitted light or reflected light of the light, a driving means for moving or rotating the inspection object, and the imaging for every predetermined movement amount or every predetermined rotation amount of the inspection object The image data of the predetermined area cut out by the cut-out means and the cut-out means cut out a predetermined area around the portion representing the transmitted light or the reflected light from each of the plurality of picked-up images picked up by the means are combined and combined And a composing means for generating an appearance surface image representing the state of the appearance surface of the inspection object.

  According to the appearance surface image generating apparatus according to the present invention, the inspection object is moved or rotated by the driving means. A predetermined area around a portion representing transmitted light or reflected light is cut out from each of a plurality of captured images picked up by the image pickup means for each predetermined movement amount or predetermined rotation amount of the inspection object by the cutout means.

  Then, the synthesizing unit arranges and synthesizes the image data of the predetermined area cut out by the clipping unit, and generates an appearance surface image representing the state of the appearance surface of the inspection object.

  In this way, by cutting out a predetermined area around a portion representing transmitted light or reflected light from each of a plurality of captured images and arranging and combining the image data of the cut out predetermined areas, the inspection can be performed with a simple configuration. An image representing the state of the appearance surface of the object can be generated.

  The appearance surface image generation apparatus according to the present invention can further include defect detection means for detecting, as a defect portion, a portion having a predetermined luminance or higher from the appearance surface image generated by the combining means. Thereby, it is possible to detect a defective portion of the appearance surface of the inspection object with a simple configuration.

  The inspection object according to the present invention is a metal member or resin member having a rotationally symmetric shape, the imaging unit images a region including reflected light of linear light from the appearance surface of the inspection object, and the driving unit rotates. The object to be inspected can be rotated with the symmetrical rotation center axis as the rotation axis.

  The inspection object of the present invention is a plate-shaped transparent member, the imaging means images an area including transmitted light of linear light from the appearance surface of the inspection object, and the driving means is an inspection object, It can be moved in a direction parallel to the plate-like surface.

  When the inspection object is a metal member or resin member having a rotationally symmetric shape, the clipping means represents reflected light from pixels whose luminance is not more than a predetermined value for each line orthogonal to the rotation axis of the captured image. A predetermined region can be cut out by cutting out a predetermined number of pixels opposite to the portion.

  In addition, the cutout unit can cut out a predetermined region by cutting out a predetermined number of pixels determined for each line based on the radius or diameter of the inspection object and a predetermined rotation amount.

  In addition, the above clipping unit determines the position of the pixel to be cut out and the predetermined number of pixels to be cut out by obtaining a moving average with the preceding and succeeding lines for the position of the pixel to be cut out and the predetermined number of pixels to be cut out for each line. be able to.

  When the inspection object is a plate-shaped transparent member, the clipping means specifies a portion representing transmitted light for each of a plurality of captured images, expands the specified portion, and among the expanded regions An area that does not represent transmitted light can be cut out as a predetermined area from the captured image.

  As described above, according to the appearance surface image generation device of the present invention, a predetermined region around a portion representing transmitted light or reflected light is cut out from each of a plurality of captured images, and an image of the cut out predetermined region is obtained. By arranging the data side by side, it is possible to generate an image representing the state of the appearance surface of the inspection object with a simple configuration.

It is a block diagram which shows the structure of the external appearance inspection system which concerns on the 1st Embodiment of this invention. It is an image figure which shows a mode that the test target object, area camera, line illumination, and back illumination are arrange | positioned. It is an image figure which shows a mode that the angle of reflected light changes with surfaces. It is a graph which shows the change of the brightness | luminance of a pixel line. (A) It is an image figure which shows the image which imaged the test target object, (B) It is an image figure which shows an external surface image. It is a flowchart which shows the content of the defect detection process routine in the external appearance inspection system which concerns on the 1st Embodiment of this invention. It is a flowchart which shows the content of the captured image acquisition process routine in the external appearance inspection system which concerns on the 1st Embodiment of this invention. (A) The figure which shows the example of R average image, (B) The figure which shows the example of G average image, (C) The figure which shows the example of the screen which shows an external surface image and a defect candidate. It is an image figure which shows a mode that the test object, the drive stage, the area camera, and the line illumination were arrange | positioned. It is an image figure which shows the image which imaged the test target object. It is a figure for demonstrating the pitch between frames at the time of synthesize | combining a cut-out area | region. It is a flowchart which shows the content of the defect detection process routine in the external appearance inspection system which concerns on the 2nd Embodiment of this invention. (A) The figure which shows the example of the image which imaged the lens which is a test object, (B) The figure which shows the example of the screen which shows an external surface image and a defect candidate. (A) The figure which shows the example of the image which imaged the touchscreen which is a test object, (B) The figure which shows the example of the screen which shows an external surface image and a defect candidate.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the present embodiment, a case where the present invention is applied to an appearance surface inspection system that detects a three-dimensional defect from an appearance surface of a rotationally symmetric inspection object will be described as an example.

  As shown in FIG. 1, the appearance inspection system 10 according to the first embodiment rotates a rotationally symmetric (for example, cylindrical) metal member, which is an inspection object, about a rotationally symmetric rotation center axis. The driving unit 12 that rotates as an axis, the line illumination 14 that irradiates the green linear slit light to the external surface of the inspection object, and the red light from the opposite side of the line illumination 14 to the inspection object A back illumination 16 that irradiates light, an area camera 18 that captures an area including light reflected from the linear illumination light from the line illumination 14, and light from the back illumination 16, and a captured image captured by the area camera 18. Based on this, an external surface image representing the state of the external surface of the inspection object is generated, and a computer 20 that detects defects on the external surface and displays them on the display device 22 is provided.

  As shown in FIG. 2, the line illumination 14 is arranged so that the linear slit light parallel to the rotation axis of the inspection object 24 is irradiated on the appearance surface of the inspection object 24, and the linear slit light An area camera 18 is arranged so that an area including reflected light is imaged. Further, the back illumination 16 and the area camera 18 are arranged so that light from the back illumination 16 is imaged through the inspection object 24. The line illumination 14, the back illumination 16, and the area camera 18 are fixed, and the inspection object 24 is installed on the drive stage of the drive unit 12 so that only the inspection object 24 is rotated.

  Depending on the inspection object to be imaged, the imaging direction angle, illumination type, illumination method, and illumination angle of the camera where defects such as scratches and dacon are shining best differ for each type of inspection object. The angle in the imaging direction of the camera, the type of illumination, how to apply, and the angle are obtained in advance, and the area camera 18, the line illumination 14, and the back illumination 16 are installed according to the type of inspection object.

  The area camera 18 includes a plurality of light receiving elements that output signals corresponding to visible light incident on each of a plurality of pixels arranged two-dimensionally, and converts signals output from the plurality of light receiving elements into digital signals. Then, a color image is generated.

  The line illumination 14 includes a plurality of LEDs arranged in a line in the length direction of the linear slit light to be irradiated and a cylindrical lens that is long in the LED arrangement direction, and light emitted from the plurality of LEDs is emitted from the LEDs. Irradiated as a linear slit light through a cylindrical lens.

  The computer 20 includes a CPU, a ROM that stores a program of a defect detection processing routine that will be described later, a RAM that stores data, and a bus that connects these. If the computer 20 is described by functional blocks divided for each function realizing means determined based on hardware and software, as shown in FIG. 1, the computer 20 rotates the inspection object 24 by the drive unit 12. A drive control unit 28 for controlling, an image acquisition unit 30 for controlling the imaging by the area camera 18 according to the rotation of the inspection object 24 and acquiring a plurality of color images continuously output from the area camera 18; An image processing unit 32 that performs image processing on the color image to generate an appearance surface image representing the unevenness of the appearance surface of the inspection object 24, and detects a three-dimensional defect candidate from the appearance surface image. And a defect detection unit 34 for outputting the appearance surface image and the detection result to the display device 22.

The drive control unit 28 controls the drive unit 12 to rotate the inspection object by a predetermined amount of rotation every time the area camera 18 performs imaging. Note that the number of frames to be imaged in one rotation of the inspection object is determined, and the rotation amount (rotation angle) is determined in advance from the number of frames according to the following equation (1).
Rotation amount a [deg / frame] = 360 ° / number of frames (1)

  The image acquisition unit 30 controls to perform imaging by the area camera 18 when driving by the driving unit 12 is stopped. The image acquisition unit 30 includes, for example, an A / D converter and an image memory that stores image data, and acquires and stores a plurality of color images output from the area camera 18.

  Next, the principle of the embodiment according to the present invention will be described.

  As described above, the line illumination 14 and the area camera 18 are installed, and images are continuously taken while the drive stage on which the inspection object is installed is moved a certain distance.

  Here, as shown in FIGS. 3A to 3C, the angle at which the reflected light enters the area camera varies depending on the surface. Therefore, in the obtained image data, the boundary portion where light and dark are separated by line illumination is the portion that is most easily confirmed with respect to a three-dimensional defect (such as a scratch or a dacon). Extracting the part where the light and dark are separated from the continuously captured images as slit images, and joining the slit images together to generate a plain image, thereby obtaining an image representing the uneven state of the appearance surface of the inspection object Can do.

  Therefore, in the present embodiment, as described below, the image processing unit 32 generates an appearance surface image based on a plurality of color images.

  First, for each acquired color image, the image is rotated so that the y-axis direction becomes the rotation axis direction. For example, when the rotation axis is the x-axis direction in the acquired color image, the image is rotated 270 ° so that the vertical axis (y-axis) direction is the rotation axis direction.

  For each pixel of the color image, the color is reduced from 24 bits to 8 bits of R and G, and converted into an R image and a G image.

  Next, smoothing is performed in order to minimize the shake of the rotation axis, and an image with reduced noise is generated. For example, from 100 frames of R images obtained from all 100 frames of captured images, a plurality of frames are selected at predetermined frame intervals, such as 0th frame, 10th frame, 20th frame, 30th frame, etc. An average of the R images is taken to generate an R average image. Similarly, a G average image is also generated.

  Next, as will be described below, the cutout start position and cutout length are determined for each line that is in the direction orthogonal to the rotation axis of the image.

  First, the luminance value of each line is examined from the R average image and the G average image. For example, the same line is cut out from each of the R average image and the G average image, and those with high luminance are preferentially overlapped to extract a change in luminance value of the line as shown in FIG.

Further, based on the change in the luminance with respect to the X direction, the portion where the luminance value has fallen from the average value of the luminance from the portion representing the reflected light (high luminance portion) is specified as the cutout start position Xs. Also, two portions where the luminance is increased by the back illumination 16 are extracted, and the diameter or radius r of the inspection object is specified from the interval between the extracted portions. Then, using the rotation amount (rotation angle) a determined above, the cutout length is determined according to the following equation (2).
Cutout length Length [number of pixels] = 2 * π * r * a / 360 (2)

After determining the cut start position and cut length for all lines, the change of the cut start position determined for each line is smoothed by a moving average with the preceding and following lines. For example, according to the following equation (3), the moving average Xs ′ with the preceding and following lines is calculated for the cutout start position of the target line.
Xs ′ = (X (s−1) + 2Xs + X (s + 1)) / 4 (3)

  However, Xs is the cutout start position of the target line, X (s−1) is the cutout start position of the line immediately before the target line, and X (s + 1) is one after the target line. This is the line cut start position.

Further, according to the following equation (4), the moving average Ls ′ with the preceding and succeeding lines is calculated for the cutout length of the target line.
Ls ′ = (L (s−1) + 2Ls + L (s + 1)) / 4 (4)

  However, Ls is the cut-out length of the target line, L (s-1) is the cut-out length of the line immediately before the target line, and L (s + 1) is the line after the target line. The cut length.

  Note that the moving average width may be set by a parameter.

  As described above, the two values determined for each line (cutout start position Xs ′, cutout length Length) are held.

  Further, the cutout start position is arbitrarily adjusted for each line. Based on the cutout start position determined as described above, the cutout start position Xs ′ + α is determined for each line s using the adjustment parameter α set in advance to delay the start position.

  Then, for each line, the cut start position and cut length determined as described above are read, and are specified from the cut start position and cut length of each line as shown in FIG. 5A from all captured images. A region (region consisting of a pixel row of each line from the cutout start position to a position away from the portion representing the reflected light by a cutout length) is cut out. The cut out regions are connected so as to be arranged for all the captured images, and as shown in FIG. 5B, a plane image (plane image) in which a rotationally symmetric shape is developed is generated as an appearance surface image.

  As described above, the image processing unit 32 performs an image process to generate an appearance surface image representing the uneven state of the entire appearance surface of the inspection object. The area specified from the cutout start position and the cutout length is an example of a predetermined area around the portion representing the reflected light of the present invention.

  The defect detection unit 34 detects, as a defect candidate, a portion where the luminance value is greater than or equal to the specified luminance (a value that can be adjusted with parameters) from the generated appearance surface image. Further, the defect detection unit 34 classifies the defect for each detected defect candidate based on the brightness, size, and shape. For example, when the brightness is low, it is classified as possible dirt or fingerprints, and when the brightness is large and the shape is round, it is classified as possible dacon or bubbles.

  Next, the operation of the appearance surface inspection system 10 according to the present embodiment will be described. As an inspection object, a rotationally symmetric metal member is placed on the drive stage. Further, when the operator inputs and sets the number of frames to be imaged during one rotation of the inspection object with respect to the computer 20, the rotation amount is determined according to the above equation (1). Further, the computer 20 executes a defect detection processing routine shown in FIG.

  First, in step 100, a plurality of captured images are acquired. Step 100 is realized by the captured image acquisition processing routine shown in FIG.

  In step 120, a variable n for identifying the frame number of the image to be captured is set to an initial value of 1. In step 122, control is performed so that the area camera 18 performs imaging, and the image is output from the area camera 18. Get a color image.

  In the next step 124, the color image acquired in step 122 is stored in a memory (not shown) as a bitmap (BMP) file.

  In step 126, the drive unit 12 is controlled to rotate the inspection object by a predetermined amount of rotation, and the inspection object is rotated by the amount of rotation. In the next step 128, it is determined whether or not the variable n is less than a constant N representing the number of frames to be imaged during one rotation. If the variable n is less than the constant N, in step 130, The variable n is incremented and the process returns to step 122. On the other hand, when a color image is captured for the number of frames to be captured during one rotation and the variable n reaches a constant N, the captured image acquisition processing routine is terminated.

  Then, in step 102 of the defect detection processing routine, each of the color images acquired in step 100 is converted into an R image and a G image, and the R image selected at a predetermined frame interval is shown in FIG. Such an R average image is generated. Further, a G average image as shown in FIG. 8B is generated from the G image selected at a predetermined frame interval.

  In the next step 104, the change of the luminance value of each line is extracted from the R average image and the G average image generated in step 102, the cut start position and cut length are determined for each line, and the cut region is determined. decide.

  In step 106, the cutout area determined in step 104 is read from each of all the color images acquired in step 100. In step 108, the image data of the cutout area read in step 106 are arranged and combined. By doing so, an appearance surface image is generated.

  In step 110, a portion having a luminance value equal to or higher than a predetermined value is detected as a defect candidate from the appearance surface image generated in step 108, and a defect is classified for each defect candidate.

  In the next step 112, the appearance surface image generated in step 108 and the defect candidate detected in step 110 are output to the display device 22, and a screen as shown in FIG. Then, the defect detection processing routine is terminated.

  As described above, according to the appearance inspection system according to the first embodiment, the area specified by the cutout start position and the cutout length around the portion representing the reflected light from each of the plurality of captured images. By cutting out and combining the image data of the cut out regions side by side, it is possible to generate an appearance surface image representing the uneven state of the appearance surface of the inspection object with a simple configuration. Further, by detecting a region where the luminance is greater than or equal to a predetermined value, it is possible to detect a defective portion on the appearance surface of the inspection object with a simple configuration.

  In the above-described embodiment, the case where the position where the luminance value has fallen below the average value is determined as the extraction start position from the change in luminance in the pixel line has been described as an example. However, the present invention is not limited to this. Instead, the cutout start position may be determined by another method. For example, a position of n% (can be set by a parameter) from the diameter of the inspection target may be determined as the cut start position.

  In addition, the case where the rotationally symmetrical metal member is used as the inspection object has been described as an example, but the present invention is not limited thereto, and a rotationally symmetrical member having light reflectivity may be used as the inspection object. For example, a resin member or a transparent member may be used as the inspection object.

  Next, a second embodiment will be described. In addition, since the structure of the external appearance inspection system which concerns on 2nd Embodiment is the same as that of 1st Embodiment, description is abbreviate | omitted.

  In the second embodiment, the point that the plate-shaped transparent member is the inspection object and the point that the transmitted light of the linear slit light from the inspection object is imaged by the area camera are the first implementation. It is mainly different from the form.

  In the appearance inspection system according to the second embodiment, the inspection object is a plate-like transparent member, and for example, a lens or a touch panel is the inspection object. In addition, a plate-shaped transparent member shall also contain the lens of the shape formed by a plane and a part spherical shape.

  The drive unit 12 moves the plate-shaped transparent member that is the inspection object in the plane axis direction (direction parallel to the surface of the inspection object). For example, as shown in FIG. 9, the inspection object 224 is installed on a transparent drive stage of the drive unit 12, and the inspection object 224 is moved by moving the drive stage in the plane axis direction of the inspection object 224. Moving.

  Further, the line illumination 14 is arranged so as to irradiate the external surface of the inspection object 224 with linear slit light parallel to the plane of the inspection object 224, and an area including the transmitted light of the linear slit light is imaged. The area camera 18 is arranged to do so. The line illumination 14 and the area camera 18 are fixed.

  The drive control unit 28 controls the drive unit 12 so as to move the inspection object by an amount determined in advance each time an image is captured by the area camera 18.

  The image acquisition unit 30 controls to perform imaging by the area camera 18 when driving by the driving unit 12 is stopped. The image acquisition unit 30 acquires and stores a color image output from the area camera 18. Note that the number of frames to be imaged is determined based on the length of the inspection object to be imaged and the amount of movement between the arbitrarily determined images. Moreover, it is preferable to adjust the area camera 18 so that the pixel size is a multiple of the moving resolution.

  As described below, the image processing unit 32 generates an appearance surface image based on a plurality of color images.

  First, the following processing is performed for each of a plurality of color images.

  From the target color image, a high-luminance portion that is equal to or higher than a predetermined luminance value is extracted, and a labeling process is performed to identify the largest region among the extracted high-luminance portions. This specified region is determined as a line portion which is a portion representing the transmitted light of the linear slit light. Further, for the smoothing process, as shown in FIG. 10, the line portion is expanded by a specified amount, and the expanded region is determined as the line portion.

  Next, the line portion determined as described above is expanded by a predetermined amount to determine a region of interest ROI (Region of Interest). At this time, the range of the region of interest ROI is expanded to a portion where valid image data is obtained. Then, as shown in FIG. 10, a region obtained by removing the above-described line portion from the region of interest ROI is determined as a cut-out region, which is a region effective for inspection, that is, a portion where unevenness is most clearly understood. This cutout area becomes an inspection area.

  With the above processing, a cutout region is determined for each of the plurality of color images. The cut-out area is an example of a predetermined area around the portion representing the transmitted light according to the present invention.

Next, a corresponding cutout region is cut out from each of all the color images. The image data of the cut-out regions that have been cut out are synthesized so as to overlap each other with a predetermined pitch shift for all color images as shown in FIG. It should be noted that a higher luminance is given priority to the pixels to be overlapped when combining. Further, the pitch between frames at the time of combining is determined according to the following expression (5) based on the amount of parameter movement at the time of imaging and the pixel size.
Pitch between frames = movement amount [um] / pixel size [um] (5)

  As described above, the image processing unit 32 performs the above-described image processing to generate an appearance surface image representing the uneven state of the entire appearance surface of the inspection target.

  Next, a defect detection processing routine according to the second embodiment will be described with reference to FIG. In addition, about the process similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted. Further, a case where a lens is used as an inspection object will be described as an example.

  First, in step 250, a plurality of captured images are acquired. Step 250 is realized by a captured image acquisition process routine.

  In the captured image acquisition routine according to the present embodiment, first, control is performed so as to perform imaging by the area camera 18, and a color image as shown in FIG. 13A output from the area camera 18 is acquired and acquired. The color image is stored in a memory (not shown) as a bitmap (BMP) file.

  Then, the drive unit 12 is controlled to move the inspection object by a predetermined movement amount, and the inspection object is moved by the movement amount.

  The above processing is repeated until a color image is captured for the number of frames to be captured, and the captured image acquisition processing routine is terminated.

  In step 252 of the captured image acquisition processing routine, a line portion representing transmitted light of the linear slit light is determined from each of the color images acquired in step 250.

  In the next step 254, the line portion of each color image determined in the above step 252 is expanded by a predetermined amount to determine each region of interest. From the region of interest determined in step 254, the cut-out region is determined by excluding the line portion determined in step 252.

  In step 258, the corresponding cutout area determined in step 256 is read from each of all the color images acquired in step 250. In step 260, the image data of the cutout area read in step 258 is read. Are synthesized while shifting by a predetermined pitch to generate an appearance surface image.

  In step 110, a portion having a luminance value equal to or higher than a predetermined value is detected as a defect candidate from the appearance surface image generated in step 260, and a defect is classified for each defect candidate.

  In the next step 112, the appearance surface image generated in step 260 and the defect candidates detected in step 110 are output to the display device 22, and a screen as shown in FIG. Then, the defect detection processing routine is terminated.

  When a touch panel is used as the inspection object, a color image as shown in FIG. 14A is captured by executing the defect detection processing routine, and the generated appearance surface image and the detected defect are detected. The candidates are output to the display device 22 and a screen as shown in FIG. 14B is displayed on the display device 22.

  As described above, according to the appearance inspection system according to the second embodiment, the portion representing the transmitted light of the linear slit light excluding the line portion from the region of interest from each of the plurality of captured images. By cutting out peripheral cutout areas and arranging the image data of the cutout areas side by side, it is possible to generate an external appearance image representing the uneven state of the external appearance surface of the inspection object with a simple configuration. Further, by detecting a region where the luminance is greater than or equal to a predetermined value, it is possible to detect a defective portion on the appearance surface of the inspection object with a simple configuration.

  Next, a third embodiment will be described. In addition, since the structure of the external appearance inspection system which concerns on 3rd Embodiment is the same as that of 1st Embodiment, description is abbreviate | omitted.

  In the third embodiment, the point that the plate-shaped resin member is the inspection object and the point that the reflected light of the linear slit light from the inspection object is imaged by the area camera are the second embodiment. It is mainly different from the form.

  In the appearance inspection system according to the third embodiment, the inspection object is a plate-shaped resin member.

  The drive part 12 moves the plate-shaped resin member which is a test object in the plane axis direction. Further, the line illumination 14 is arranged so as to irradiate the external surface of the inspection object 224 with linear slit light parallel to the plane of the inspection object 224, and an area including the reflected light of the linear slit light is imaged. The area camera 18 is arranged to do so.

  As described below, the image processing unit 32 generates an appearance surface image based on a plurality of color images.

  First, the following processing is performed for each of a plurality of color images.

  From the target color image, a high-luminance portion that is equal to or higher than a predetermined luminance value is extracted, and a labeling process is performed to identify the largest region among the extracted high-luminance portions. This specified region is determined as a line portion that is a portion representing the reflected light of the linear slit light. Further, for the smoothing process, the line portion is expanded by a specified amount, and the expanded region is determined as the line portion.

  Next, the region of interest ROI is determined by expanding the line portion determined as described above by a predetermined amount. And the area | region remove | excluding said line part from the region of interest ROI is determined as a cutting-out area | region.

  With the above processing, a cutout region is determined for each of the plurality of color images. The cut-out area is an example of a predetermined area around the portion representing the reflected light of the present invention.

  Next, a corresponding cutout region is cut out from each of all the color images. The cut-out image data of the cut-out area is synthesized so as to be overlapped with respect to all color images while shifting by a predetermined pitch to generate an appearance surface image. It should be noted that a higher luminance is given priority to the pixels to be overlapped when combining.

  As described above, the image processing unit 32 performs the above-described image processing to generate an appearance surface image representing the uneven state of the entire appearance surface of the inspection target.

  In addition, about the other structure and effect | action of an external appearance inspection system which concern on 3rd Embodiment, since it is the same as that of 2nd Embodiment, description is abbreviate | omitted.

  In this way, from each of the plurality of captured images, the cut-out region around the portion representing the reflected light of the linear slit light, excluding the line portion from the region of interest, is cut out, and the image data of the cut-out regions are arranged side by side and synthesized. By doing so, it is possible to generate an appearance surface image representing the uneven state of the appearance surface of the inspection object with a simple configuration.

  In the above embodiment, the case where the plate-shaped resin member is used as the inspection object has been described as an example. However, the present invention is not limited to this, and a plate-shaped member having light reflectivity is used as the inspection object. For example, a metal member or a transparent member may be used as the inspection object.

  In the first to third embodiments described above, the case where imaging is performed by an area camera every time the inspection object is moved by a certain amount has been described as an example. However, the present invention is not limited to this. Instead, for example, images may be captured continuously while moving the drive stage to obtain a predetermined number of captured images. In this case, referring to the movement amount (for example, the length of the inspection object), the speed of the drive stage and the frame rate of the area camera are calculated, and the inspection object is moved at the calculated predetermined speed, Continuous imaging is performed at the calculated predetermined frame rate. When the number of frames to be imaged is captured, the imaging is terminated and the movement is terminated, and each captured image obtained is saved as a bitmap file.

DESCRIPTION OF SYMBOLS 10 Appearance surface inspection system 12 Drive part 14 Line illumination 16 Back illumination 18 Area camera 20 Computer 22 Display apparatus 24, 224 Inspection object 28 Drive control part 30 Image acquisition part 32 Image processing part 34 Defect detection part

Claims (10)

A light irradiation means for irradiating the external surface of the inspection object with linear light;
Imaging means for imaging a region including transmitted light or reflected light of the linear light from the appearance surface of the inspection object;
Drive means for moving or rotating the inspection object;
Cutting out a predetermined region around the portion representing the transmitted light or the reflected light from each of a plurality of captured images captured by the imaging unit for each predetermined movement amount or predetermined rotation amount of the inspection object Means,
Combining the image data of the predetermined area cut out by the cut-out means to synthesize and generate an appearance surface image representing the state of the appearance surface of the inspection object;
An external surface image generation apparatus including:   The appearance surface image generation apparatus according to claim 1, further comprising defect detection means for detecting a portion having a predetermined luminance or higher as a defect portion from the appearance surface image generated by the synthesizing means. The inspection object is a rotationally symmetric member having light reflectivity,
The imaging means images a region including reflected light of the linear light from the appearance surface of the inspection object,
The appearance surface image generation apparatus according to claim 1, wherein the driving unit rotates the inspection object about a rotation center axis of the rotationally symmetric shape as a rotation axis. The inspection object is a plate-shaped transparent member,
The imaging means images an area including transmitted light of the linear light from the appearance surface of the inspection object,
The appearance surface image generation apparatus according to claim 1, wherein the driving unit moves the inspection object in a direction parallel to a plate-like surface. The inspection object is a plate-like member having light reflectivity,
The imaging means images a region including reflected light of the linear light from the appearance surface of the inspection object,
The appearance surface image generation apparatus according to claim 1, wherein the driving unit moves the inspection object in a direction parallel to a plate-like surface.   The cutout means cuts out the predetermined number of pixels on the opposite side of the portion representing the reflected light from pixels whose luminance is equal to or lower than a predetermined value for each line orthogonal to the rotation axis of the captured image. The appearance surface image generation apparatus according to claim 3, wherein the region is cut out.   The appearance according to claim 6, wherein the cutout unit cuts out the predetermined region by cutting out the predetermined number of pixels determined for each line based on a radius or a diameter of the inspection object and the predetermined rotation amount. Surface image generation device.   The cutout means determines the position of the pixel to be cut out and the predetermined number of pixels to be cut out by obtaining a moving average of the positions of the cutout pixels and the predetermined number of cutout pixels with respect to the preceding and succeeding lines for each line. Item 8. The appearance surface image generation device according to Item 7.   The cutout unit specifies a portion representing the transmitted light for each of the plurality of captured images, expands the specified portion, and selects a region that is not a portion representing the transmitted light among the expanded regions. The external surface image generation apparatus according to claim 4, wherein the external image is cut out from the captured image as the predetermined region.   The clipping means identifies a portion representing the reflected light for each of the plurality of captured images, expands the identified portion, and selects a region that is not a portion representing the reflected light among the expanded regions. 6. The appearance surface image generation apparatus according to claim 5, wherein the image is cut out as the predetermined area from the captured image.