A method for detecting defects in stamping parts
By locking the rectangular detection area and standard model through machine vision equipment, adjusting the probe position and image rotation, the problem of inaccurate high-definition image acquisition caused by the offset of stamping parts is solved, and the accuracy and comprehensiveness of the detection of surface defects of stamping parts are improved.
Patent Information
- Application Number
- CN202411224851.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-03
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-09-03
AI Technical Summary
In the prior art, the position offset of the stamping member during transmission causes the probe to be unable to accurately acquire high-definition images, affecting the accuracy of image comparison.
The rectangular detection area and standard model are determined by machine vision equipment, the center point of the stamping member is locked, the probe position and image rotation are adjusted, the stamping member images are centeredly aligned within the standard detection area, and area comparison is performed to generate detection signals.
Effectively obtain high-definition images of stamping parts, ensure accurate position alignment, and improve the accuracy and comprehensiveness of surface defect detection of stamping parts.
Smart Images

Figure CN118735921B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of stamping parts, and in particular to a method for detecting defects in stamping parts. Background Art
[0002] Stamping is a plastic processing process that uses die pressure to deform the sheet metal to obtain parts with a certain shape, size and performance. The parts obtained after stamping are called stamping parts.
[0003] Application Publication No. CN117830139B discloses a method for defect detection of high-strength stamping parts. The method obtains the comprehensive grayscale eigenvalues of pixels in a stamping part image; obtains the edge significance of the pixels in the original stamping part grayscale image based on the difference between the distribution of the comprehensive grayscale eigenvalues of the pixels in the original stamping part grayscale image and the corresponding matching pixels in the sub-stamping part grayscale image at each scale; uses this to adjust the preset grayscale value standard deviation, and enhances the original stamping part grayscale image based on the adjusted grayscale value standard deviation to obtain an enhanced grayscale image of the stamping part for defect detection. The present invention adaptively adjusts the preset grayscale value standard deviation based on the edge significance, suppresses noise points in the image, improves image denoising efficiency, and increases the accuracy of high-strength stamping part defect detection.
[0004] During the relevant inspection process, the stamping parts are located on a designated conveyor belt for transmission, and in the initial stage of transmission, the working oscillation relationship of certain working components will cause the position of the stamping parts on the conveyor belt to shift. Since the changed position of the stamping parts is not determined, the acquisition direction of the corresponding probe cannot be determined, which will lead to differences in the clarity of the acquired stamping parts images. As a result, the image comparison accuracy of some stamping parts needs to be strengthened during the subsequent image comparison. Summary of the Invention
[0005] In response to the shortcomings of the existing technology, the present invention provides a method for detecting defects in stamping parts, which solves the problem that the original probe cannot move relative to the specific position of the stamping part to ensure the accuracy of image comparison.
[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for detecting defects in stamping parts, comprising the following steps:
[0007] Step 1: Based on the location of the machine vision device, confirm the relevant area monitored by this vision device. Based on the straight line contour of the conveyor belt edge, lock the rectangular area within the relevant area and fill the set standard model into this rectangular area to obtain the standard detection area. The specific sub-steps are as follows:
[0008] S11. Based on the location of the machine vision device, determine a relevant area monitored by the machine vision device, where the relevant area is located on the conveyor belt;
[0009] S12. Generate a set of rectangular areas from the relevant area, where the edge lines of both sides of the rectangular areas coincide with the edge lines of both ends of the conveyor belt. Generate diagonals of the rectangular areas based on the relevant diagonal points of the rectangular areas. Then, determine the intersection of the diagonals and mark it within the rectangular areas. This intersection is the center point of the rectangular areas.
[0010] S13. Determine the center point of a pre-set standard model. This center point is pre-set within the standard model. Based on the center point of the standard model and the center point of the rectangular area, place the standard model in the middle of the rectangular area to obtain a standard detection area. The placement direction of the standard model is pre-set.
[0011] Step 2: Use machine vision equipment to acquire images of stamping parts entering the standard inspection area. Based on the center point of the stamping part image, identify whether the position of the stamping part meets the standard. Based on the identification result, assess whether the internal probe of the machine vision equipment needs to be moved, and control the internal probe to move to the specified position to acquire a high-definition image of the stamping part surface. The specific sub-steps are as follows:
[0012] S21. When a stamping part enters a standard inspection area, identify whether the contour of the stamping part intersects the standard inspection area. If so, continue monitoring until the contour of the stamping part does not intersect the standard inspection area. Determine the edge contour of the stamping part based on the surface image of the stamping part, place the edge contour in a two-dimensional coordinate system, determine the point coordinates of several points within the contour line of the edge contour, average the coordinates of the several points to determine a set of mean coordinates, and determine the mean point based on the mean coordinates. Synchronously calibrate within the area included in the edge contour to determine the center point position of the stamping part.
[0013] S22. Construct a set of standard vertical lines that pass through the center point of the standard inspection area and are perpendicular to the side lines of the standard inspection area. Identify whether the center point of the stamping part is within the standard vertical lines:
[0014] If not, construct a set of defined perpendicular lines passing through the center point of the stamping part and perpendicular to the two side lines of the standard inspection area, determine the perpendicular point where the center point inside the standard inspection area is perpendicular to the defined perpendicular lines, control the internal probe to move from the center point of the stamping part to the position of the perpendicular point, and when the center point of the stamping part moves to coincide with the perpendicular point, control the internal probe to obtain a high-definition image of the surface of the stamping part;
[0015] If so, when the center point of the stamping part moves to coincide with the center point inside the standard inspection area, the internal probe is controlled to obtain a high-definition image of the surface of the stamping part;
[0016] Step 3: Based on the high-definition image of the stamping part surface taken by the machine vision equipment, the confirmed high-definition image is compared with the standard inspection area, and the rotation position of the high-definition image is adjusted. The specific sub-steps are:
[0017] S31. Based on the determined standard detection area, a standard model is set in the middle of the standard detection area. Based on the four corner points and the center point of the standard detection area, four dividing lines are determined by connecting the corner points and the center point. The four dividing lines divide the standard detection area into four areas. Then, the standard area of the upper surface of the standard model in each area is determined in turn and calibrated as BZ. i , where i represents different regions, and i=1, 2, 3, 4;
[0018] S32. Identify whether the center point of the high-definition image coincides with the center point of the standard detection area:
[0019] If they overlap, directly confirm whether the area of the high-definition image in the four areas is consistent with the BZ of the corresponding area. i If they are consistent, no processing is performed. If they are inconsistent, the HD image is directly rotated, and the rotation point is the center point of the HD image, until the areas of the four regions determined in the corresponding rotation process are consistent with the BZ of the corresponding region. i Stop when they are consistent and determine the current position;
[0020] If they do not coincide, move the HD image until the center of the HD image coincides with the center of the standard detection area, then rotate the HD image in such a way that the centers of the two coincide, so that the areas of the four regions determined in the HD image coincide with the BZ of the corresponding regions. i consistent;
[0021] Step 4: For the HD image after correct position adjustment, compare different areas in the HD image with the upper surface image of the standard model in the standard detection area to identify whether the surface comparison results of the two are consistent. Based on the identification results, determine whether to generate a detection abnormality signal. The specific sub-steps are:
[0022] S41. Based on the four areas divided in the standard detection area, the standard areas BZ of the four areas are preferentially identified. i Are all consistent? If all are consistent, go to step S411. If there are two-way consistency or two groups of consistency, go to step S421. If not all are consistent, go to step S431.
[0023] S411. Compare the high-definition image with the upper surface image of the standard model in the standard detection area to determine whether the comparison results are consistent. If they are consistent, generate a normal detection signal. If they are inconsistent, rotate the high-definition image 90 degrees clockwise or counterclockwise based on its calibrated center point to determine whether the comparison results are consistent:
[0024] If it is still inconsistent, continue to rotate 90 degrees until the high-definition image is rotated to the initial position and still inconsistent, then directly generate a detection abnormality signal and display it;
[0025] If they are consistent, a detection normal signal is generated;
[0026] S421. Compare the high-definition image with the upper surface image of the standard model in the standard detection area to determine whether the comparison results are consistent. If they are consistent, generate a normal detection signal. If they are inconsistent, rotate the high-definition image 180 degrees clockwise or counterclockwise based on its calibrated center point to determine whether the comparison results are consistent:
[0027] If there is still inconsistency, a detection anomaly signal is directly generated and displayed;
[0028] If they are consistent, a detection normal signal is generated;
[0029] S431. Directly compare the high-definition image with the upper surface image of the standard model in the standard detection area: if they are consistent, a normal detection signal is generated; if they are inconsistent, an abnormal detection signal is generated and displayed.
[0030] The present invention provides a method for detecting defects in stamping parts. Compared with the prior art, it has the following advantages:
[0031] The present invention defines the relevant detection area, and based on the real-time acquired surface image of the stamping part, preferentially identifies whether the center point of the stamping part is aligned with the midpoint of the detection area, and then adjusts the high-definition image of the stamping part based on the specific area parameters of the standard model located in the corresponding detection area. This not only effectively obtains the high-definition image of the stamping part, but also simultaneously moves the high-definition image to the center position of the detection area to ensure the accurate position alignment of the stamping part, facilitates subsequent related detection, and can effectively detect surface defects of stamping parts at different positions.
[0032] When inspecting defects on the surface of stamping parts, different surface comparison methods are performed based on the area parameters of each different area within the inspection area to determine whether the stamping parts need to be rotated in a related manner, thereby making the surface defect inspection of stamping parts more comprehensive and achieving better surface defect detection results. BRIEF DESCRIPTION OF THE DRAWINGS
[0033] Figure 1 Schematic diagram of the process of the present invention;
[0034] Figure 2 Schematic diagram for determining the rectangular area of the present invention. DETAILED DESCRIPTION
[0035] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.
[0036] Example 1
[0037] See also Figure 1 , the present application provides a stamping part defect detection method, comprising the following steps:
[0038] Step 1. Based on the location of the machine vision equipment, confirm the relevant area monitored by this vision equipment, lock the rectangular area from the relevant area based on the edge straight line contour of the conveyor belt, and fill the set standard model into this rectangular area to obtain the standard detection area. Specifically, the machine vision equipment is provided with a monitoring probe to acquire images of the detection area, and when the stamping parts are inspected, the stamping parts produced in succession are generally placed in the middle position of the conveyor belt by the relevant manipulator. After the stamping parts are moved to the designated position through the transmission effect of the conveyor belt, the surface high-definition image of the stamping parts is acquired through the monitoring probe, and then the defect detection, comparison and identification are performed on such surface high-definition images. If the stamping parts shake due to the operation of other devices, resulting in the stamping parts being offset from the center position, then the clarity of the relevant high-definition images acquired by the monitoring probe may have some errors, which will lead to inaccurate detection.
[0039] The specific sub-steps of locking the detection area are as follows:
[0040] S11. Combination Figure 2 , based on the location of the machine vision device, determine the relevant area monitored by the machine vision device (after the vision device is installed and adjusted, it will monitor the designated area, and this designated area is the relevant area), and this relevant area is located on the conveyor belt;
[0041] S12. Generate a set of rectangular areas within the relevant area, where the edge lines of both sides of the rectangular areas coincide with the edge lines of both ends of the conveyor belt. Generate diagonals of the rectangular areas based on the relevant diagonal points of the rectangular areas. Then, determine the intersection of the diagonals and mark it within the rectangular areas. This intersection is the center point of the rectangular areas. The high-definition probe inside the machine vision device is also aligned with this center point to obtain a high-definition image.
[0042] S13. Determine the center point of a pre-set standard model (this standard model is a standard comparison model corresponding to the stamping part, which is pre-set by the relevant operator). This center point is pre-set in the standard model. Based on the center point of the standard model and the center point of the rectangular area, place the standard model in the rectangular area, and make the standard model in the middle of the rectangular area to obtain a standard test area. The placement direction of the standard model is pre-set, generally, the central axis of the standard model passes through the center point of the rectangular area and is perpendicular to the upper and lower sides of the rectangular area.
[0043] Step 2: Use machine vision equipment to acquire images of stamping parts entering the standard inspection area. Based on the center point of the stamping part image, identify whether the position of the stamping part meets the standard. Based on the identification result, assess whether the internal probe of the machine vision equipment needs to be moved, and control the internal probe to move to the specified position to acquire a high-definition image of the stamping part surface.
[0044] The specific sub-steps for identifying whether the position of the stamping part meets the standards are as follows:
[0045] S21. When a stamped part enters a standard inspection area, identify whether the contour of the stamped part intersects the standard inspection area. If so, continue monitoring until the contour of the stamped part does not intersect the standard inspection area (that is, the stamped part completely enters the standard inspection area due to the action of the conveyor belt). Determine the edge contour of the stamped part based on the surface image of the stamped part, place the edge contour in a two-dimensional coordinate system, determine the point coordinates of several points within the contour line of the edge contour, average the coordinates of the several points to determine a set of mean coordinates, and determine the mean point based on the mean coordinates. Synchronously calibrate the area included in the edge contour to determine the center point position of the stamped part.
[0046] S22. Construct a set of standard vertical lines that pass through the center point of the standard inspection area and are perpendicular to the side lines of the standard inspection area. Identify whether the center point of the stamping part is within the standard vertical lines:
[0047] If so, when the center point of the stamping part moves to coincide with the center point inside the standard inspection area, the internal probe is controlled to obtain a high-definition image of the surface of the stamping part;
[0048] If not, construct a set of defined perpendicular lines passing through the center point of the stamping part and perpendicular to the two side lines of the standard inspection area, determine the perpendicular point where the center point inside the standard inspection area is perpendicular to the defined perpendicular lines, control the internal probe to move from the center point of the stamping part to the position of the perpendicular point, and when the center point of the stamping part moves to coincide with the perpendicular point, control the internal probe to obtain a high-definition image of the surface of the stamping part;
[0049] Specifically, during the transmission of stamped parts, if the initial position of the conveyor belt is close to the relevant working device, the device will vibrate during operation, causing the stamped parts that are placed in the correct position to shift on the conveyor belt. Then, after such stamped parts shift, in the subsequent image recognition process, because the position captured by the original camera has not changed, there will be a corresponding difference in the clarity of the captured image, which will cause errors in the detection process.
[0050] In order to make the captured image clear enough, it is necessary to adjust the facing position of the corresponding camera in real time based on the changes in the center point of the stamping part to ensure that the captured image of the stamping part is clear.
[0051] Step 3: Based on the high-definition image of the stamping part surface captured by the machine vision device, the confirmed high-definition image is compared with the standard inspection area, and the rotation position of the high-definition image is adjusted. The specific sub-steps for adjustment are:
[0052] S31. Based on the determined standard detection area, a standard model is set in the middle of the standard detection area. Based on the four corner points and the center point of the standard detection area, four dividing lines are determined by connecting the corner points and the center point. The four dividing lines divide the standard detection area into four areas. Then, the standard area of the upper surface of the standard model in each area is determined in turn and calibrated as BZ. i , where i represents different regions, and i=1, 2, 3, 4;
[0053] S32. Identify whether the center point of the high-definition image coincides with the center point of the standard detection area:
[0054] If they overlap, directly confirm whether the area of the high-definition image in the four areas is consistent with the BZ of the corresponding area. i If they are consistent, no processing is performed. If they are inconsistent, the HD image is directly rotated, and the rotation point is the center point of the HD image, until the areas of the four regions determined in the corresponding rotation process are consistent with the BZ of the corresponding region. i Stop when they are consistent and determine the current position;
[0055] If they do not coincide, move the HD image until the center of the HD image coincides with the center of the standard detection area, then rotate the HD image in such a way that the centers of the two coincide, so that the areas of the four regions determined in the HD image coincide with the BZ of the corresponding regions. i consistent;
[0056] The reason for the relevant rotation here is to ensure that the two images are oriented in the same direction when the images are compared later, so as to ensure the relevant accuracy in the subsequent detection process, to determine whether the comparison between the corresponding images is consistent, and thus to identify whether there are relevant defects on the surface of the stamping part and display the signal;
[0057] Example: There is a high-definition image whose center point does not coincide with the center point of the standard detection area. First, the system will move the image so that the two center points coincide. Then, the system will check whether the area of the image in the four zones coincides with the pre-set BZ. i If they are not consistent, the system will rotate the image and continuously detect the areas of the four regions during the rotation until these areas are consistent with BZ i This process ensures that the image is accurately positioned and oriented before further analysis or processing.
[0058] Step 4: For the HD image after correct position adjustment, compare different areas in the HD image with the upper surface image of the standard model in the standard detection area to identify whether the surface comparison results of the two are consistent. Based on the identification results, determine whether to generate a detection abnormality signal. The specific sub-steps of the determination are as follows:
[0059] S41. Based on the four areas divided in the standard detection area, the standard areas BZ of the four areas are preferentially identified. i Are all consistent? If all are consistent, go to step S411. If there are two-way consistency or two groups of consistency, go to step S421. If not all are consistent, go to step S431.
[0060] S411. Compare the high-definition image with the upper surface image of the standard model in the standard detection area to determine whether the comparison results are consistent. If they are consistent, generate a normal detection signal. If they are inconsistent, rotate the high-definition image 90 degrees clockwise or counterclockwise based on its calibrated center point (that is, the calibrated center point in the high-definition image) to determine whether the comparison results are consistent:
[0061] If it is still inconsistent, continue to rotate 90 degrees until the high-definition image is rotated to the initial position and still inconsistent, then directly generate a detection abnormality signal and display it;
[0062] If they are consistent, a detection normal signal is generated;
[0063] For example, consider a high-definition image with four regions whose standard areas, BZ_i and BZi, are all identical. First, the system compares this image with an image of the top surface of a standard model. If the comparison indicates an inconsistency, the system rotates the image 90° and compares again until a consistent angle is found or all possible angles are confirmed to be inconsistent, generating a corresponding signal. This process ensures the correct orientation of the image before further analysis or processing.
[0064] S421. Compare the high-definition image with the upper surface image of the standard model in the standard detection area to determine whether the comparison results are consistent. If they are consistent, generate a normal detection signal. If they are inconsistent, rotate the high-definition image 180 degrees clockwise or counterclockwise based on its calibrated center point (that is, the calibrated center point in the high-definition image) to determine whether the comparison results are consistent:
[0065] If there is still inconsistency, a detection anomaly signal is directly generated and displayed;
[0066] If they are consistent, a detection normal signal is generated;
[0067] S431. Directly compare the high-definition image with the upper surface image of the standard model in the standard detection area: if they are consistent, a normal detection signal is generated; if they are inconsistent, an abnormal detection signal is generated and displayed;
[0068] Specifically, if the corresponding stamping part is circular or square, its surrounding area will be consistent. Then, after the center point and related area of its high-definition image are adjusted, the stamping part may still be inconsistent with the standard model. If the end of the stamping part is facing down and the end of its standard model is facing up, if there are corresponding mounting holes or other forms of surface grooves on the end, then the comparison result will show inconsistency. However, there is actually no problem with such stamping parts. Therefore, the angle change method is implemented here to perform relevant comparisons to ensure the accuracy and comprehensiveness of the detection process.
[0069] Similarly, if it is a rectangle, the areas of its two ends also correspond one to one, and then it belongs to the specific situation disclosed in step S421.
[0070] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.
[0071] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.
Claims
1. A stamping part defect detection method, characterized in that: The following steps are involved: Step 1: Based on the location of the machine vision device, confirm the relevant area monitored by the vision device, lock the rectangular area from the relevant area based on the edge straight line contour of the conveyor belt, and fill the set standard model into the rectangular area to obtain the standard detection area. The specific sub-steps are: S11. Based on the location of the machine vision device, determine a relevant area monitored by the machine vision device, where the relevant area is located on the conveyor belt; S12, generating a set of rectangular areas from the relevant area, wherein the edge lines on both sides of the rectangular area coincide with the edge lines on both ends of the conveyor belt, generating diagonal lines based on the relevant diagonal points of the rectangular area, and then determining the intersection of the diagonal lines and marking them in the rectangular area, wherein the intersection point is the center point of the rectangular area; S13, determining the center point of a preset standard model, the center point being preset in the standard model, placing the standard model in the rectangular area based on the center point of the standard model and the center point of the rectangular area, and making the standard model in the middle of the rectangular area, to obtain a standard detection area, and the placement direction of the standard model is preset in advance; Step 2: Use machine vision equipment to obtain images of stamping parts entering the standard inspection area. Based on the center point of the stamping part image, identify whether the position of the stamping part meets the standard. Based on the identification result, assess whether the internal probe of the machine vision equipment needs to be moved, and control the internal probe to move to the specified position to obtain a high-definition image of the stamping part surface. The specific sub-steps are: S21. When the stamping part enters the standard inspection area, identify whether the contour of the stamping part intersects with the standard inspection area. If it does, continue monitoring until the contour of the stamping part does not intersect with the standard inspection area, and then determine the edge contour of the stamping part based on the surface image of the stamping part. Then, place the edge contour in a two-dimensional coordinate system, determine the point coordinates of several points inside the contour line of the edge contour, average the coordinates of several points, determine a set of mean coordinates, and determine the mean point based on the mean coordinates, and synchronously calibrate in the area included in the edge contour to determine the center point position of the stamping part; S22. Construct a set of standard vertical lines that pass through the center point of the standard inspection area and are perpendicular to the side lines of the standard inspection area, and identify whether the center point of the stamping part is within the standard vertical lines: If not, construct a set of defined vertical lines that pass through the center point of the stamping part and are perpendicular to the side lines of the standard inspection area on both sides, determine the vertical point of the vertical line that is perpendicular to the center point inside the standard inspection area and the defined vertical line, control the internal probe to move from the center point of the stamping part to the position of the vertical point, and when the center point of the stamping part moves to coincide with the vertical point, control the internal probe to obtain a high-definition image of the surface of the stamping part; If the center point of the stamping part is within the standard vertical line, when the center point of the stamping part moves to coincide with the center point inside the standard inspection area, the internal probe is controlled to obtain a high-definition image of the surface of the stamping part; Step 3: Based on the high-definition image of the stamping surface taken by the machine vision equipment, the confirmed high-definition image is compared with the standard inspection area, and the rotation position of the high-definition image is adjusted. The specific sub-steps are: S31. Based on the determined standard detection area, a standard model is set in the middle of the standard detection area. Based on the four corner points and the center point of the standard detection area, four dividing lines are determined by connecting the corner points and the center point. The four dividing lines divide the standard detection area into four areas. Then, the standard area of the upper surface of the standard model in each area is confirmed in turn and calibrated as BZ. i , where i represents different regions, and i=1, 2, 3, 4; S32. Identify whether the center point of the high-definition image coincides with the center point of the standard detection area: If they overlap, directly confirm whether the area of the high-definition image in the four areas is consistent with the BZ of the corresponding area. i If they are consistent, no processing is performed. If they are inconsistent, the HD image is directly rotated, and the rotation point is the center point of the HD image, until the areas of the four regions determined in the corresponding rotation process are consistent with the BZ of the corresponding region. i Stop when they are consistent and determine the current position; If they do not overlap, move the HD image until the center of the HD image overlaps with the center of the standard detection area, and then rotate the HD image in such a way that the center points of the two overlap, so that the areas of the four areas determined in the HD image overlap with the BZ of the corresponding areas. i Consistency; Step 4: For the HD image after the position adjustment is correct, the different areas in the HD image are compared with the upper surface image of the standard model in the standard detection area to identify whether the surface comparison results of the two are consistent. Based on the recognition results, it is determined whether to generate a detection abnormality signal. The specific sub-steps are: S41, based on the four areas divided in the standard detection area, prioritize the identification of the standard areas BZ of the four areas i Are they all consistent? If all are consistent, execute step S411. If there are two-to-two consistency or two groups of consistency, execute step S421. If not all are consistent, execute step S431. S411, compare the high-definition image with the upper surface image of the standard model in the standard detection area to identify whether the comparison results are consistent. If they are consistent, a normal detection signal is generated. If they are inconsistent, the high-definition image is rotated 90° clockwise or counterclockwise based on its calibrated center point to identify whether the comparison results are consistent: If it is still inconsistent, continue to rotate 90° until the high-definition image is rotated to the initial position and still inconsistent, then directly generate a detection abnormality signal and display it; If they are consistent, a normal detection signal is generated; S421, compare the high-definition image with the upper surface image of the standard model in the standard detection area to identify whether the comparison results are consistent. If they are consistent, a normal detection signal is generated. If they are inconsistent, the high-definition image is rotated 180° clockwise or counterclockwise based on its calibrated center point to identify whether the comparison results are consistent: If it is still inconsistent, a detection anomaly signal is directly generated and displayed; If they are consistent, a normal detection signal is generated; S431. Directly compare the high-definition image with the upper surface image of the standard model in the standard detection area: if they are consistent, a normal detection signal is generated; if they are inconsistent, an abnormal detection signal is generated and displayed.
