JP4358064B2 - Laminated ring inspection method - Google Patents

Description

  The present invention relates to a method for inspecting a laminated ring formed by laminating a plurality of endless metal rings used for a belt or the like of a continuously variable transmission for a vehicle.

  A belt for a continuously variable transmission includes a large number of elements formed in a predetermined shape and a ring member that binds these elements in an annular shape. The ring member is generally composed of a laminated ring formed by laminating a plurality of endless thin metal rings. Here, the metal ring may have a scratched portion such as a dent on the side edge during a manufacturing process, a conveying process, a laminating process, or the like.

  Therefore, conventionally, prior to the metal ring laminating process, it is known that light is applied to the side edge of a single-layer metal ring, the side edge is imaged by a camera, and a scratch on the side edge of the metal ring is detected from the image. (For example, refer to Patent Document 1).

However, in this conventional example, only a single-layer metal ring is inspected, so even if a flaw is generated on the side edge of the metal ring in the lamination process, this cannot be detected, and dust is generated between the metal rings in the lamination process. Even if a stacking fault such as a foreign matter such as a bite is caught, this cannot be detected.
Japanese Patent Laying-Open No. 2004-77425 (0012, 0016, FIGS. 3 to 5)

  This invention makes it the subject to provide the inspection method of the lamination | stacking ring which enabled it to detect the lamination | stacking defect in addition to the damage | wound part of the side edge of each metal ring in view of the above point.

  In order to solve the above problems, the present invention is a method for inspecting a laminated ring formed by laminating a plurality of endless metal rings, wherein an imaging means is used with a side edge of the laminated ring illuminated by an illuminating means. An imaging process for imaging, a binarization process for binarizing the image of the imaged side edge of the laminated ring, and each metal ring based on each ring image corresponding to each metal ring appearing in the binarized image A side edge inspection process for inspecting the state of the side edges of the metal ring, and a lamination inspection process for inspecting the lamination state of the metal rings based on the relative positional relationship between the plurality of ring images.

  According to the present invention, in the side edge inspection step, it is possible to detect a flaw on the side edge of each metal ring from each ring image. In addition, when a stacking failure such as the inclusion of foreign matter occurs in the metal ring stacking process, the interval between adjacent ring images changes. Therefore, in the stacking inspection process, it is possible to detect a stacking failure of the metal rings based on the relative positional relationship between the ring images.

  Here, the stacking inspection step sequentially selects two adjacent ring images from the binarized image, the direction matching the circumferential direction of the metal ring is the X-axis direction, and the radial direction of the metal ring The Y-axis direction distance between the centroids of the two selected ring images and the Y-axis direction distance between the centroids at each position in the X-axis direction of the two selected ring images, with the matching direction as the Y-axis direction It is preferable to include a step of comparing one or more of the change amounts in the X-axis direction with respective threshold values to determine the lamination state of the metal rings.

  The side edge of the metal ring is chamfered in a rounded cross section, and the reflected light from the side edge of the metal ring entering the imaging means gradually decreases as the distance from the central portion in the thickness direction of the side edge increases. Therefore, the boundary on both sides in the Y-axis direction of the ring image before binarization does not appear clearly, and the positions of the bright and dark boundary lines (ring image edges) on both sides in the Y-axis direction of the ring image appearing in the binarized image are Y Some variation in the axial direction. For this reason, when the distance between the inner edges of two adjacent ring images is measured, even if the actual gap between the metal rings is normal, the inner edges of each ring image are displaced so that they are relatively far away. The distance between the inner edges of the ring image may be greater than or equal to a threshold value. In addition, the Y-axis direction width of the ring image that appears in the binarized image changes due to the shift of the metal ring in the axial direction, and the distance between the inner edges of the two ring images may be greater than or equal to the threshold value. . On the other hand, the center of gravity of the ring image of the binarized image (the center of the width of the ring image in the Y-axis direction) substantially coincides with the center of the metal ring in the thickness direction. Therefore, the distance between the centroids of the two ring images of the binarized image (the Y-axis direction distance between the centroids) is a parameter that accurately represents the gap amount between the two metal rings. It is possible to accurately determine the presence or absence of stacking faults due to foreign object biting or the like. In addition, the amount of gap between the metal rings may change locally due to bending of the metal rings or the entry of minute foreign objects between the metal rings, but the distance between the centers of gravity of the two ring images in the X-axis direction Such a local change in the gap amount between the metal rings can also be detected based on the change amount in.

  Further, the side edge inspection step sequentially selects one ring image from the binarized image, the direction matching the circumferential direction of the metal ring matches the X-axis direction, and the radial direction of the metal ring. A step of measuring one or more of the width, area, center of gravity, and number of changes in brightness in the Y-axis direction at each position in the X-axis direction for the selected ring image with the direction as the Y-axis direction. And the deviation between the ring image Y-axis direction width and the average width, the ring image Y-axis direction width change in the X-axis direction, the ring image area minimum value, and the ring image area change amount in the X-axis direction. It is desirable to provide a step of determining the state of the side edge of the metal ring by comparing one or more of the amount of change in the X-axis direction of the ring image in the X-axis direction and the number of light-dark changes with the respective threshold values.

  Here, the size and shape of the scratches that may occur on the side edges of the metal ring are various, and the scratches that appear in the image also have various forms, which are overlooked in uniform image processing such as pattern matching. There will also be scratched parts. As described above, the deviation between the Y-axis direction width and the average width of the ring image, the amount of change in the Y-axis direction width of the ring image in the X-axis direction, the minimum value of the ring image area, and the change in the ring image area in the X-axis direction By detecting various items such as the amount, the amount of change in the X-axis direction of the center of gravity of the ring image, and the number of light and dark changes, it is possible to prevent detection of a flaw.

  As a binarization method, in order to prevent the shape and size of the image appearing in the binarized image from changing greatly due to variations in the luminance of the captured image, a bright portion having a higher luminance portion as a bright portion. Two threshold values, a threshold value for darkness and a threshold value for dark portion where the lower luminance portion is a dark portion, are used, and the luminance portion between the bright portion threshold value and the dark portion threshold value becomes light gray corresponding to the luminance. There is a method of binarizing as described above. In this case, it is appropriate to label the gray part of the binarized image as a bright part and label the continuous bright part as one ring image. Then, when measuring the area of the ring image in the side edge inspection process, if the area of the gray portion included in the ring image is calculated as a value obtained by multiplying the actual area by a coefficient corresponding to the brightness of the gray portion, When the ring image contains a gray scratch image corresponding to a hollow-shaped scratch portion with a shallow side edge, the ring image area becomes smaller at the scratch image portion, and the minimum ring image area or the ring image is reduced. The presence of a flaw image can be detected based on the amount of change in the X-axis direction of the area.

  Referring to FIG. 1, reference numeral 1 denotes a rotary table. On this rotary table 1, a laminated ring W ′ formed by laminating a plurality of endless metal rings W (12 in the illustrated example) as shown in FIG. It can be set freely. Each metal ring W is laminated with a slight gap with the adjacent metal ring W. Moreover, the side edge of each metal ring W is chamfered in a cross-sectional round shape. Above the turntable 1, an imaging unit composed of the imaging means 2 and the illumination means 3 is disposed so as to face the side edge of the laminated ring W ′. Further, an image processing device 4 for processing an image picked up by the image pickup means 2 and a rotary table control device 5 for controlling the rotation of the rotary table 1 are provided. The rotary table 1, the imaging unit, and image processing are provided. The device 4 and the rotary table control device 5 constitute an inspection device for the laminated ring W ′.

  Although not shown, the rotary table 1 is provided with a holder that contacts the inner peripheral surface of the laminated ring W ′ and holds the laminated ring W ′ in an annular shape. Further, on the side of the turntable 1, a marking device 6 having a marker 6a that advances and retreats toward the laminated ring W ′ is provided.

  The imaging means 2 is composed of a digital camera such as a CMOS camera, and is connected to the image processing device 4 via a cable. The illumination means 3 includes a ring-shaped irradiation head 31 equipped with a light diffusion filter 31a, and a light source 33 connected to the irradiation head 31 via an optical fiber 32, and light from the light source 33 is transmitted to the ring-shaped irradiation head. Irradiation from 31 toward the side edge of the laminated ring W ′. The imaging unit 2 images the side edge of the laminated ring W ′ through the inner circumferential space of the ring-shaped irradiation head 31.

  As shown in FIG. 2, the image processing apparatus 4 includes, as a functional configuration, an image storage unit 41, a binarization unit 42, a side edge inspection unit 43, a stacking inspection unit 44, a result display unit 45, and a position notification unit 46. I have. The result display means 45 is connected to a monitor 47 provided in the image processing apparatus 4, and the position notification means 46 is connected to the turntable control apparatus 5.

  The rotary table control means 5 includes a marking control means 51 connected to the marking device 6. Then, when a scratched portion or a stacking failure is detected in the metal ring W, the stacked ring W ′ in which a scratched portion or the like exists due to the rotation of the rotary table 1 based on the position information of the scratched portion transmitted from the position notification means 46. The marking device 6 is controlled so that the marker 6a operates when this portion faces the marking device 6, and marking is performed on the portion where the flaws and the like of the laminated ring W ′ exist.

  Next, the inspection procedure for the laminated ring W ′ will be described. When inspecting the laminated ring W ′, first, the rotary table 1 is rotated at a predetermined speed by the rotary table control device 5 while the laminated ring W ′ is placed on the rotary table 1, and the imaging means 2 detects the laminated ring W ′. The side edges are continuously imaged (STEP 1 in FIG. 4). For example, when the imaging is a laminated ring W ′ in which a plurality of metal rings W having a standard circumference of 615 mm are laminated, 6 mm of the circumference of the laminated ring W ′ is regarded as one image, and each image has a lapping margin of 1 mm. Thus, the entire circumference is set to 123 divided images.

  Imaging by the imaging unit 2 is performed in a state where the side edge of the laminated ring W ′ is illuminated by the illumination unit 3. As a result, a plurality of bright strip-shaped images corresponding to the plurality of metal rings W constituting the laminated ring W ′ appear in the image of the imaging means 2. Since the side edge of the metal ring W is chamfered with a rounded cross section as described above, the light incident on the side portion in the thickness direction of the side edge of the metal ring W is not reflected to the imaging means 2 side. The part between the bright band-like images becomes dark. The images of the imaging unit 2 are sequentially sent to the image processing device 4 and stored in the image storage unit 41.

  Next, the image stored in the image storage means 41 is binarized by the binarization means 42 (STEP 2 in FIG. 4). In this binarization processing, a bright portion (white) threshold value (threshold value for making a portion of luminance higher than this white by binarization) and a dark portion (black) threshold value (a luminance portion less than this value are binarized). The image is binarized so that the luminance portion between the light portion threshold value and the dark portion threshold value becomes gray of lightness corresponding to the luminance. In the binarized image, each ring image w corresponding to the side edge of each metal ring W appears as a bright portion as shown in FIG. In addition, although the gray part also appears in the binarized image, in FIG. 5, for the sake of convenience, the bright part including the gray part is represented by white and the dark part (black) is represented by shading. Further, the binarizing means 42 performs a labeling process on the ring image w appearing in the binarized image. In the labeling process, each ring image w is labeled with one gray image as a group of consecutive bright portions, considering the gray portion as a bright portion. In FIG. 6 described later, the shaded portion represents a dark portion unless otherwise specified.

  Next, a side edge inspection process for inspecting the state of the side edge of each metal ring W based on the ring image w is performed by the side edge inspection means 43. In the side edge inspection process, first, one ring image w is selected from a plurality of labeled ring images w in a predetermined order (STEP 3 in FIG. 4). The length of the selected ring image w in the X-axis direction is defined as a direction that matches the circumferential direction of the metal ring W in the binarized image as an X-axis direction and a direction that matches the radial direction of the metal ring W as a Y-axis direction. Is measured (STEP 4 in FIG. 4), and it is determined whether or not the length is less than a predetermined threshold (STEP 5 in FIG. 4). As shown in FIG. 6A, the ring image is divided into a ring image w of # 1 and a ring image w of # 2 due to cracks or scratches on the metal ring W, and the length of the ring image w in the X-axis direction is shortened. In this case, it is rejected (STEP 32 in FIG. 4). Note that the length in the X-axis direction of the ring image w of # 3 in FIG. 6A is a normal length.

  If the length of the ring image w in the X-axis direction is equal to or greater than a predetermined threshold, then one image is divided into a plurality of equal widths (for example, one pixel width) in the X-axis direction, and the ring image in each division is obtained. Measure the area of w. In consideration of the possibility that the ring image w may include a gray portion, a lightness coefficient corresponding to the lightness is set such that white is 1 and black is 0, and the area of each portion corresponds to the lightness of the portion. The area of the ring image w is obtained by multiplying the brightness coefficient. Next, the minimum area of all the sections is obtained (STEP 6 in FIG. 4), and it is determined whether or not the minimum area of the ring image w is less than a predetermined threshold (STEP 7 in FIG. 4). If there is a scratched part such as a dent, chipping or shaving on the side edge of the metal ring W, a scar image wa which is a dark part image corresponding to the scratched part enters the ring image w as shown in FIG. The area of w becomes small and is rejected. Note that the area may be less than the threshold even in a section other than the section where the area is the lowest, but for the purpose of shortening the processing time, the areas other than the minimum area are not determined as pass / fail. Here, the inspection process is performed on the image of each divided portion obtained by dividing the entire circumference of the laminated ring W ′ into 123 as described above. The divided portion of the laminated ring W ′ corresponding to the image is marked and visually inspected in a later process. If there is a scratched portion other than the minimum area of the divided portion, the scratched portion is visually checked. Since it can be easily confirmed by inspection, there is no practical problem even if only the minimum area is set as a pass / fail discrimination target.

  If the minimum area of the ring image w is equal to or greater than a predetermined threshold value, then, the difference between the ring image area of each section and the ring image area of the section adjacent thereto is used as the amount of change in the X-axis direction of the ring image area. The maximum value of the change amount in the X-axis direction (area maximum change amount) of the ring image area in one image is obtained (STEP 8 in FIG. 4), and it is determined whether or not this area maximum change amount exceeds a predetermined threshold value. (STEP 9 in FIG. 4). When a shallow dent-like scratch exists on the side edge of the metal ring W, a gray scratch image wb corresponding to the scratch may appear in the ring image w as shown in FIG. In FIG. 6C, the dark part is represented by black and the gray part is represented by shading. Since the area is calculated by multiplying the lightness coefficient as described above in the portion of the scar image wb, the area of the ring image w decreases in the portion where the scar image wb exists, and the ring image area in the boundary portion of the scar image wb. The amount of change in the X-axis direction increases. If the maximum area change amount exceeds the threshold value, it is rejected.

  If the maximum area change amount is less than or equal to the threshold value, then the position of the center of gravity (area center of gravity) of the ring image w of each section is obtained, and the difference between the position of the center of gravity of each section and the position of the center of gravity of the section adjacent thereto is calculated. The amount of change in the X-axis direction of the center of gravity of the ring image w is calculated, and the maximum value of the amount of change in the X-axis direction of the center of gravity of the ring image w in one image (the center of gravity maximum change amount) is obtained (STEP 10 in FIG. 4). It is determined whether or not the maximum change amount exceeds a predetermined threshold (STEP 11 in FIG. 4). If the metal ring W is bent due to interference with manufacturing equipment or the like, the ring image w is bent as shown in FIG. 6D, and the center of gravity change amount is increased at the bent portion. If the maximum change in the center of gravity exceeds the threshold value, the result is rejected.

  If the maximum change in the center of gravity is equal to or less than the threshold value, the Y-axis direction width (ring image width) of the ring image w of each section is measured. In the measurement of the ring image width, the gray portion is also regarded as a bright portion, and the Y-axis direction from the lowermost bright portion and dark portion boundary in the Y-axis direction to the uppermost bright portion and dark portion boundary in the Y-axis direction. The distance is obtained and this is set as the ring image width. Then, the difference between the ring image width of each section and the ring image width of the section adjacent thereto is calculated as the amount of change in the ring image width in the X-axis direction, and the maximum amount of change in the X-axis direction of the ring image width in one image is calculated. A value (maximum change amount of width) is obtained (STEP 12 in FIG. 4), and it is determined whether or not the maximum change amount of width exceeds a predetermined threshold (STEP 13 in FIG. 4). In this case, depending on the shape of the scratch, the ring image w becomes wider as shown in FIG. 6E, and the width change amount becomes larger in this portion. If the maximum width change amount exceeds the threshold value, the result is rejected. Further, when the central portion of the scratched portion is shallow and recessed, a gray portion wc may appear in the wide portion of the ring image w as shown in FIG. In FIG. 6E, the dark part is represented by black and the gray part is represented by shading. In this case, since the area of the ring image w is multiplied by the brightness coefficient by the gray portion wc, the area does not increase as much as the left, and the scratches may be overlooked in the determination based on the above-described maximum area change amount. Therefore, performing the determination based on the maximum width change amount is useful for preventing an oversight of a flaw.

  If the maximum width change amount is less than or equal to the threshold value, then the number of light and dark changes in the Y-axis direction is measured in each of the above sections (STEP 14 in FIG. 4). When measuring the number of light-dark changes, the gray part is also considered a bright part, and each section is scanned from the upper side to the lower side in the Y-axis direction, and when the dark part changes to the bright part and when the dark part changes to the bright part. 1 is added to the counter of the number of changes in brightness. Normally, the number of light / dark changes is 2. However, if a scar image wd that is a dark part is included in the ring image w as shown in FIG. 6 (f), the number of light / dark changes when scanned along the arrow line in FIG. 4 Then, it is determined whether or not there is a section in which the number of changes in light and dark exceeds 2 which is a threshold value (STEP 15 in FIG. 4). If there is a section in which the number of changes in light and darkness is 3 or more, it is rejected. It should be noted that the count number may be set to 1 when the light portion shifts to the dark portion and further shifts to the bright portion, and may be rejected when there is a section having the count number of 1.

  If the number of light-dark changes is 2 or less, next, the average value (average width) of all the ring image widths of the respective sections is obtained (STEP 16 in FIG. 4), and the average width of the ring image widths of the respective sections is determined. A deviation from the average width (+ direction width deviation) is calculated for a large one (STEP 17 in FIG. 4), and it is determined whether or not the + direction width deviation exceeds a predetermined threshold (STEP 18 in FIG. 4). When the side edge of the metal ring W has an overhanging defect due to a handling mistake or the like, as shown in FIG. 6G, an overhanging portion we in the Y-axis direction corresponding to the overhanging defect appears in the ring image w. The + direction width deviation becomes large at the overhanging portion we. If the + direction width deviation exceeds the threshold value, it is rejected.

  If the + direction width deviation is less than or equal to the threshold value, next, a deviation from the average width (−direction width deviation) is calculated for the ring image width of each section smaller than the average width (STEP 19 in FIG. 4). It is determined whether or not the direction width deviation exceeds a predetermined threshold (STEP 20 in FIG. 4). As shown in FIG. 6H, when the wound image wa which is a dark part corresponding to the scratched part on the side edge of the metal ring W enters the ring image w, the -direction width deviation becomes large at the part of the scratched image wa. And when-direction width deviation exceeds the said threshold value, it is set as the failure.

  Here, the pass / fail judgment based on the deviation of the ring image width with respect to the average width is a measure against the deviation of the metal ring W in the axial direction. That is, when any of the metal rings W is displaced upward in the axial direction (direction approaching the image pickup means 2), the width of the ring image w corresponding to the metal ring W is widened, and the area of the ring image w becomes a scratch image. The location where wa exists is relatively large. Therefore, the determination based on the minimum area in STEP 7 may not be rejected. On the other hand, by performing determination based on the deviation of the ring image width with respect to the average width, it is possible to avoid as much as possible the presence of a flaw due to the influence of the change in the ring image width due to the shift of the metal ring W.

  -If the direction width deviation is less than or equal to the threshold value, then, after adding 1 to the ring counter that counts the number of inspected ring images w (STEP 21 in FIG. 4), has all the ring images w been inspected? It is determined whether or not (STEP 22 in FIG. 4), and the process returns to STEP 3 until the inspection of all the ring images w is completed, and the above-described side edge inspection process is repeated.

  When the inspection of all the ring images w is completed, it is determined whether or not the count number of the ring counter matches the number of metal rings W to be stacked as the stacked ring W ′ (number of rings = 12) (FIG. 4 STEP23). If any of the metal rings W is missing or the metal rings W are excessively stacked, the count number does not match the number of rings, and in this case, the count is rejected.

  If the count number matches the number of rings, the stacking inspection unit 44 performs stacking inspection processing for inspecting the stacking state of the metal rings W based on the relative positional relationship between the plurality of ring images. In the stacking inspection process, first, two ring images w and w adjacent in the Y-axis direction in a predetermined order are selected from a plurality of labeled ring images w (STEP 24 in FIG. 4). Next, the position of the center of gravity (area center of gravity) in one image of the two selected ring images w and w is obtained, and the Y axis between the center of gravity of one ring image w and the center of gravity of the other ring image w A direction distance (distance between centroids) is calculated (STEP 25 in FIG. 4), and it is determined whether or not the distance between centroids exceeds a predetermined threshold (STEP 26 in FIG. 4). When a foreign object is caught between the metal rings W and W or the metal ring W is bent, a ring image # 1 corresponding to the metal ring where the foreign object is caught or bent is shown in FIG. The distance between the centers of gravity of the ring images w of # 2 and # 2 is increased. The distance between the centers of gravity of the ring image w of # 2 and the ring image w of # 3 is a normal size. If the distance between the centroids exceeds the threshold value, the result is rejected.

  If the distance between the centroids is less than or equal to the threshold, then one image is divided into a plurality of (for example, six) with the same width in the X-axis direction, and the respective centroids (areas) of the two ring images w and w in each division. The position of the center of gravity) is obtained, and the Y-axis direction distance (distance between centers of gravity) between the center of gravity of one ring image w and the center of gravity of the other ring image w in each section is calculated (STEP 27 in FIG. 4). Then, the amount of change in the X-axis direction of the distance between the centroids is calculated as the difference between the distance between the centroids in each section and the distance between the centroids in the adjacent sections. A value (maximum change in distance between centroids) is obtained (STEP 28 in FIG. 4), and it is determined whether or not the maximum change in distance between centroids exceeds a predetermined threshold (STEP 29 in FIG. 4). When the metal ring W is bent, the amount of change in the distance between the centers of gravity increases as shown in FIG. If the maximum change in the distance between the centroids exceeds the threshold value, the determination is rejected.

  Instead of the center of gravity of two adjacent ring images w, w, the position of the light / dark boundary line inside the adjacent direction of the ring images w, w is measured, and the distance between the light / dark boundary lines of the two ring images w, w is measured. It is also possible to determine the stacking state of the metal rings W based on the Y-axis direction distance and the amount of change in this distance in the X-axis direction. However, the position of the light / dark boundary line of each ring image w in the binarized image is displaced due to variations in the brightness of the image of the imaging means 2, and the Y axis of the ring image w due to the shift of the metal ring W in the axial direction. It is displaced even if the direction width changes. As a result, even if the gap between two adjacent metal rings W, W is normal, the distance in the Y-axis direction between the light and dark boundary lines of the two ring images w, w and the amount of change in the X-axis direction of this distance. Exceeds the respective threshold values, and even if the gap between two adjacent metal rings W, W is abnormal, Y between the light and dark boundary lines of the two ring images w, w There is a possibility that the axial distance and the amount of change in this distance in the X-axis direction are less than the respective threshold values, resulting in erroneous determination as acceptable. On the other hand, the center-of-gravity position of each ring image w in the binarized image is the width of the ring image w in the Y-axis direction due to the influence of variations in the brightness of the image of the imaging means 2 and the displacement of the metal ring W in the axial direction. The metal ring W almost coincides with the center in the thickness direction without being influenced by the change of the left. Therefore, as in this embodiment, the center of gravity positions of two adjacent ring images w and w are measured, and the stacking state is based on the distance between the center of gravity of the two ring images w and w and the amount of change in the distance between the center of gravity. It is more advantageous to discriminate misidentification.

  If the maximum change in the distance between the center of gravity is less than or equal to the threshold value, it is next determined whether or not the inspection has been completed for all the combinations of the two adjacent ring images w and w (STEP 30 in FIG. 4). Until this is done, the process returns to STEP 24 and the above-described stacking inspection process is repeated. In this way, the stacking inspection process is performed on all combinations of two adjacent ring images w and w, and if it is not determined to be unacceptable at the stage where this is completed, it is accepted (STEP 31 in FIG. 4). . Then, the side edge inspection process and the lamination inspection process are performed on all images captured by dividing the laminated ring W ′, and when the inspection of any image is accepted, the laminated ring W ′ is assembled. The next process is paid out. On the other hand, if it is rejected at any stage, that fact is displayed on the monitor and marking is applied to the portion of the laminated ring W ′ corresponding to the inspected image. Then, a visual inspection of the laminated ring W ′ is performed by the inspector with the marking as a mark, and the correction work such as removing the metal ring W having a scratch or the like or removing the foreign matter caught between the metal rings is performed. Do.

  In the above-described embodiment, the area change amount, the centroid change amount, and the width change amount of the ring image w in the side edge inspection process are represented by the area of the ring image w in each section in the X-axis direction and the section adjacent thereto. Although the difference between the center of gravity position and the width is obtained, it is also possible to obtain the difference between the area, the center of gravity position, and the width of the ring image w between each section and a section away from the center by a predetermined distance in the X-axis direction. Moreover, in the said embodiment, although the ring-shaped light irradiation head 31 is used as the illumination means 3, and the imaging location by the imaging means 2 of the side edge of lamination | stacking ring W 'is made to irradiate light from all the circumference | surroundings. For example, in addition to the light from the ring-shaped irradiation head 31, the spot light may be irradiated from a direction inclined in the radial direction with respect to the axial direction of the laminated ring W ′. Further, without using the ring-shaped irradiation head 31, spot light or diffused light may be irradiated from a direction inclined in the circumferential direction with respect to the axial direction of the laminated ring W ′.

The system block diagram which shows the inspection apparatus used for implementation in this invention inspection method. The block diagram which shows the functional structure of the image processing apparatus and rotary table control apparatus of the inspection apparatus of FIG. A cut perspective view of a part of a laminated ring The flowchart which shows the test | inspection procedure of this invention test | inspection method. Explanatory drawing which shows the binarized image of the lamination | stacking ring imaged with the imaging means of the inspection apparatus of FIG. (A)-(j) Explanatory drawing which shows the various forms of the ring image which appears in a binarized image.

Explanation of symbols

  W: Metal ring, W ′: Laminated ring, w: Ring image, 2 ... Imaging means, 3 ... Illuminating means, 4 ... Image processing device.

Claims (2)

A method for inspecting a laminated ring formed by laminating a plurality of endless metal rings,
An imaging step of imaging by the imaging means in a state where the side edge of the laminated ring is illuminated by the illumination means;
A binarization step for binarizing the image of the side edge of the imaged laminated ring;
A side edge inspection step of inspecting the state of the side edge of each metal ring based on each ring image corresponding to each metal ring appearing in the binarized image;
A lamination inspection step for inspecting the lamination state of the metal ring based on the relative positional relationship between a plurality of ring images ;
The stacking inspection step sequentially selects two adjacent ring images from the binarized image, and matches the circumferential direction of the metal ring with the X-axis direction and the radial direction of the metal ring. The Y axis direction distance between the centroids of the two selected ring images and the Y axis direction distance between the centroids at each position in the X axis direction of the two selected ring images inspection method of a laminated ring, wherein Rukoto and a step of determining the stacked state of the metal ring as compared to the one or more respective threshold of the amount of change in the X-axis direction. A method for inspecting a laminated ring formed by laminating a plurality of endless metal rings,
An imaging step of imaging by the imaging means in a state where the side edge of the laminated ring is illuminated by the illumination means;
A binarization step for binarizing the image of the side edge of the imaged laminated ring;
A side edge inspection step of inspecting the state of the side edge of each metal ring based on each ring image corresponding to each metal ring appearing in the binarized image;
A lamination inspection step for inspecting the lamination state of the metal ring based on the relative positional relationship between a plurality of ring images;
In the side edge inspection step, one ring image is sequentially selected from the binarized image, and the direction matching the circumferential direction of the metal ring matches the X-axis direction and the radial direction of the metal ring. A step of measuring one or more of the width, area, center of gravity, and number of changes in brightness in the Y-axis direction at each position in the X-axis direction for the selected ring image with the direction as the Y-axis direction. And the deviation between the ring image Y-axis direction width and the average width, the ring image Y-axis direction width change in the X-axis direction, the ring image area minimum value, and the ring image area change amount in the X-axis direction. , you characterized in that it comprises the step of determining the state of the side edges of the metal ring by comparing one or more of the change amount and contrast change speed in the X-axis direction of the center of gravity of the ring image and each of the threshold inspection method of the product layer ring.

Images