A brake pad detection method and system based on a CCD camera and a storage medium

Abstract

The application relates to a brake pad detection method and system based on a CCD camera and a storage medium, the method comprising the following steps: P1. In a conveying line provided with a first CCD camera and a second CCD camera, first image data information of a brake pad is acquired based on the first CCD camera, and second image data information of the brake pad is acquired based on the second CCD camera; P2. Based on the first image data information of the brake pad and the second image data information of the brake pad, EIAJ algorithm is adopted to extract edge pixel feature data information of the brake pad, and then edge pixel data information of the brake pad is output according to an edge spread function ESF; and P3. Based on the edge pixel data information of the brake pad, SFR algorithm is adopted to obtain an MTF value corresponding to the brake pad, and a predetermined threshold value is set. The application has high algorithm execution efficiency, high system stability, high detection precision and a detection error rate below 2 / 1000.

Description

Technical Field

[0001] This invention relates to the field of brake pad testing technology, and in particular to a brake pad testing method, system, and storage medium based on a CCD camera. Background Technology

[0002] With the continuous advancement of intelligent brake pad production projects, brake pad inspection devices are being used to check the appearance quality of brake pads. CCD cameras are becoming a trend in acquiring brake pad image data. At the same time, the inspection accuracy and efficiency of brake pads have become problems that we urgently need to solve.

[0003] In the prior art, patent number (CN212041564U) discloses a brake pad inspection device. The device is equipped with a first CCD detector and a second CCD detector at the inlet and outlet, respectively, and a third CCD detector is installed above the conveyor belt between the left and right strip plate limiters to collect brake pad images. Defective brake pads are pushed into the collection chute by a pusher plate in the device, while qualified brake pads are conveyed into the collection platform by the conveyor belt for collection. However, the built-in recognition algorithm of the CCD camera in this device has low recognition accuracy and slow recognition speed, often detecting good products as defective products, and is not stable enough, thus causing problems for brake pad production. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the present invention provides a brake pad detection method, system and storage medium based on a CCD camera. The algorithm has high execution efficiency and the system is very stable. In addition, the detection accuracy is high and the detection error rate is less than 0.2%.

[0005] To achieve the above and other related objectives, the present invention provides the following technical solution:

[0006] A brake pad detection method based on a CCD camera, the method comprising:

[0007] P1. In a conveyor line equipped with a first CCD camera and a second CCD camera, first image data information of the brake pad is acquired based on the first CCD camera, and second image data information of the brake pad is acquired based on the second CCD camera.

[0008] P2. Based on the first image data information and the second image data information of the brake pad, the EIAJ algorithm is used to extract the edge pixel feature data information of the brake pad, and then the edge pixel data information of the brake pad is output according to the edge expansion function ESF.

[0009] P3. Based on the edge pixel data information of the brake pad, the SFR algorithm is used to obtain the MTF value of the corresponding brake pad. A predetermined threshold is set. If the MTF value of the brake pad corresponding to the first image data information and the second image data information of the brake pad is greater than or equal to the predetermined threshold, then there is no protrusion or defect at the edge of the brake pad, and the process proceeds to step P5 for further detection; otherwise, the process proceeds to step P4.

[0010] P4. If the MTF value of the brake pad is inconsistent with the first image data information and the second image data information of the brake pad, the brake pad has a protrusion or defect at the edge, the brake pad is unqualified, and the system will issue an alarm.

[0011] P5. A hole detection algorithm is used to detect the first image data information and the second image data information of the brake pad, and the detected brake pad image data information is output respectively. The two images are compared. Based on the contrast of the image pixels, a threshold is set. If the contrast of the current pixel is greater than the threshold, it is a hole or crack in the brake pad, and the brake pad is determined to be unqualified. The system will issue an alarm.

[0012] Furthermore, in step P3, the SFR algorithm includes:

[0013] P31. Based on the edge pixel data of the brake pad, a Hough transform is used to obtain the inclination angle of the tangent line of the brake pad edge curve. The inclination angle of the tangent line of the brake pad edge curve is then fitted using the least squares method to obtain the distribution function f(x).

[0014] Where x is the inclination angle of the tangent, θ i Here, n is the number of samples;

[0015] P32. Based on the distribution function f(x), perform a two-dimensional discrete Fourier transform to obtain a periodic signal of the inclination angle of the tangent line of the brake pad edge curve;

[0016] P33. Based on the periodic signal of the inclination angle of the tangent of the brake pad edge curve, the MTF value of the corresponding brake pad is output through normalized modulus.

[0017] Furthermore, in step P5, the hole detection algorithm includes:

[0018] P51. Preprocess the first image data information and the second image data information of the brake pad to obtain the processed first and second image data information of the brake pad;

[0019] P52. Based on the processed brake pad first image data information, match the value box to obtain the value of each pixel in the value box of the brake pad first image as Y1, and based on the processed brake pad second image data information, obtain the value of each pixel in the value box of the brake pad second image as Y2.

[0020] P53. Based on the value Y1 of each pixel within the value frame of the first image of the brake pad and the value Y2 of each pixel within the value frame of the second image of the brake pad, according to the contrast function C ij , Get the contrast of the current pixel.

[0021] Furthermore, the preprocessing involves performing gamma correction on the image.

[0022] Furthermore, the range of the first image frame of the brake pad is 1 pixel, 17*17 pixels, 25*25 pixels, and 35*35 pixels; the range of the second image frame of the brake pad corresponding to the first image frame of the brake pad is 31*31 pixels, 85*85 pixels, 125*125 pixels, and 175*175 pixels.

[0023] Furthermore, the hole detection method also includes performing binarization processing on the hole or crack, calculating the area of ​​the hole or crack, and then screening and judging according to the judgment criteria.

[0024] Furthermore, in step P2, the EIAJ algorithm includes:

[0025] P21: Based on the first image data information of the brake pad and the second image data information of the brake pad, a large number of candidate regions are generated from the first and second image data information of the brake pad, these regions are filtered, and candidate regions containing brake pad images are output.

[0026] P22: Based on the candidate regions containing the brake pad image, train a network to classify the candidate regions containing the brake pad image, extract image features, and output image feature information of different regions of the brake pad.

[0027] P23: Based on the image feature information of different regions of the brake pad, the feature data information of the edge region of the brake pad is obtained by calculating the gradient map.

[0028] To achieve the above and other related objectives, the present invention also provides a brake pad detection system based on a CCD camera, including a computer device programmed or configured to perform the steps of any of the CCD camera-based brake pad detection methods described above.

[0029] To achieve the above and other related objectives, the present invention also provides a computer-readable storage medium storing a computer program programmed or configured to perform any of the CCD camera-based brake pad detection methods described herein.

[0030] The present invention has the following positive effects:

[0031] 1. This invention can accurately identify brake pad quality using only two top-view CCD cameras. By using a hole detection algorithm, the brake pad image data acquired by the first and second CCD cameras are detected, and images with holes or cracks on the brake pad surface are removed, thereby improving detection accuracy and reducing the detection error rate.

[0032] 2. This invention uses the SFR algorithm to output the corresponding brake pad MTF value, and judges the brake pad MTF value to determine whether there are protrusions or defects on the edge of the brake pad, which improves the accuracy of the algorithm, and the algorithm has high execution efficiency and high system stability.

[0033] 3. This invention reduces manual intervention, lowers labor costs, and improves the production efficiency of brake pads by combining a CCD camera with an improved algorithm. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0035] Figure 2 This is a schematic diagram of the SFR algorithm flow of the present invention;

[0036] Figure 3 This is a schematic diagram of the hole detection algorithm of the present invention;

[0037] Figure 4 This is a schematic diagram illustrating the hole detection principle of the present invention;

[0038] Figure 5 This is a schematic diagram of the production line of the present invention.

[0039] The labels in the diagram are as follows: 1—First CCD camera, 2—Second CCD camera, 3—Brake pad. Detailed Implementation

[0040] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0041] Example 1: As Figure 1 or Figure 5 As shown, a brake pad detection method based on a CCD camera is described, the method comprising:

[0042] P1. In a conveyor line equipped with a first CCD camera and a second CCD camera, first image data information of the brake pad is acquired based on the first CCD camera, and second image data information of the brake pad is acquired based on the second CCD camera.

[0043] P2. Based on the first image data information and the second image data information of the brake pad, the EIAJ algorithm is used to extract the edge pixel feature data information of the brake pad, and then the edge pixel data information of the brake pad is output according to the edge expansion function ESF.

[0044] P3. Based on the edge pixel data information of the brake pad, the SFR algorithm is used to obtain the MTF value of the corresponding brake pad. A predetermined threshold is set. If the MTF value of the brake pad corresponding to the first image data information and the second image data information of the brake pad is greater than or equal to the predetermined threshold, then there is no protrusion or defect at the edge of the brake pad, and the process proceeds to step P5 for further detection; otherwise, the process proceeds to step P4.

[0045] P4. If the MTF value of the brake pad is inconsistent with the first image data information and the second image data information of the brake pad, the brake pad has a protrusion or defect at the edge, the brake pad is unqualified, and the system will issue an alarm.

[0046] P5. A hole detection algorithm is used to detect the first image data information and the second image data information of the brake pad, and the detected brake pad image data information is output respectively. The two images are compared. Based on the contrast of the image pixels, a threshold is set. If the contrast of the current pixel is greater than the threshold, it is a hole or crack in the brake pad, and the brake pad is determined to be unqualified. The system will issue an alarm.

[0047] like Figure 2 As shown, in this embodiment, in step P3, the SFR algorithm includes:

[0048] P31. Based on the edge pixel data of the brake pad, a Hough transform is used to obtain the inclination angle of the tangent line of the brake pad edge curve. The inclination angle of the tangent line of the brake pad edge curve is then fitted using the least squares method to obtain the distribution function f(x).

[0049] Where x is the inclination angle of the tangent, θ i Here, n is the number of samples;

[0050] P32. Based on the distribution function f(x), perform a two-dimensional discrete Fourier transform to obtain a periodic signal of the inclination angle of the tangent line of the brake pad edge curve;

[0051] P33. Based on the periodic signal of the inclination angle of the tangent of the brake pad edge curve, the MTF value of the corresponding brake pad is output through normalized modulus.

[0052] In this embodiment, in step P2, the EIAJ algorithm includes:

[0053] P21: Based on the first image data information and the second image data information of the brake pad, a large number of candidate regions are generated from the first and second image data information of the brake pad, these regions are filtered, and candidate regions containing brake pad images are output.

[0054] P22: Based on the candidate regions containing the brake pad image, train a network to classify the candidate regions containing the brake pad image, extract image features, and output image feature information of different regions of the brake pad.

[0055] P23: Based on the image feature information of different regions of the brake pad, the feature data information of the edge region of the brake pad is obtained by calculating the gradient map.

[0056] In the above-mentioned brake pad inspection, the edges of the brake pads are inspected to determine whether there are any protrusions or defects, thereby determining whether the brake pads are qualified.

[0057] Example 2: Based on the brake pad detection method based on a CCD camera in Example 1, the present invention will be further described below.

[0058] like Figure 3 or Figure 4 As shown, under the condition that the brake pad edge is qualified, the brake pad surface is then inspected using a hole detection algorithm, which includes:

[0059] P51. Preprocess the first image data information and the second image data information of the brake pad to obtain the processed first and second image data information of the brake pad;

[0060] P52. Based on the processed brake pad first image data information, match the value box to obtain the value of each pixel in the value box of the brake pad first image as Y1, and based on the processed brake pad second image data information, obtain the value of each pixel in the value box of the brake pad second image as Y2.

[0061] P53. Based on the value Y1 of each pixel within the value frame of the first image of the brake pad and the value Y2 of each pixel within the value frame of the second image of the brake pad, according to the contrast function C ij , Get the contrast of the current pixel.

[0062] In this embodiment, the preprocessing is to perform gamma correction on the image.

[0063] In this embodiment, the range of the first image frame of the brake pad is 1 pixel, 17*17 pixels, 25*25 pixels, and 35*35 pixels; the range of the second image frame of the brake pad corresponding to the first image frame of the brake pad is 31*31 pixels, 85*85 pixels, 125*125 pixels, and 175*175 pixels.

[0064] In this embodiment, the hole detection method further includes performing binarization processing on the hole or crack, calculating the area of ​​the hole or crack, and then screening and judging according to the judgment criteria.

[0065] After verifying that the edge of the brake pad is qualified, the surface of the brake pad is then inspected for holes or cracks, and finally the system determines that the brake pad is qualified.

[0066] In this embodiment, the present invention also provides a brake pad detection system based on a CCD camera, including a computer device that is programmed or configured to perform the steps of any of the CCD camera-based brake pad detection methods described above.

[0067] In this embodiment, the present invention also provides a computer-readable storage medium storing a computer program programmed or configured to perform any of the CCD camera-based brake pad detection methods described above.

[0068] Any references to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0069] In summary, this invention not only has high algorithm execution efficiency and a very stable system, but also high detection accuracy, with a detection error rate of less than 0.2%.

[0070] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A brake pad detection method based on a CCD camera, characterized in that, The method includes: P1. In a conveyor line equipped with a first CCD camera and a second CCD camera, first image data information of the brake pad is acquired based on the first CCD camera, and second image data information of the brake pad is acquired based on the second CCD camera. P2. Based on the first image data information and the second image data information of the brake pad, the EIAJ algorithm is used to extract the edge pixel feature data information of the brake pad, and then the edge pixel data information of the brake pad is output according to the edge expansion function ESF. P3. Based on the edge pixel data information of the brake pad, the SFR algorithm is used to obtain the MTF value of the corresponding brake pad. A predetermined threshold is set. If the MTF value of the brake pad corresponding to the first image data information and the second image data information of the brake pad is greater than or equal to the predetermined threshold, then there is no protrusion or defect at the edge of the brake pad, and the process proceeds to step P5 for further detection; otherwise, the process proceeds to step P4. P4. If the MTF value of the brake pad is inconsistent with the first image data information and the second image data information of the brake pad, the brake pad has a protrusion or defect at the edge, the brake pad is unqualified, and the system will issue an alarm. P5. The first image data information and the second image data information of the brake pad are detected by the hole detection algorithm. The detected brake pad image data information is output respectively. The two images are compared. Based on the contrast of the image pixels, a threshold is set. If the contrast of the current pixel is greater than the threshold, it is a hole or crack in the brake pad. The brake pad is judged to be unqualified and the system will issue an alarm. In step P3, the SFR algorithm includes: P31. Based on the edge pixel data of the brake pad, a Hough transform is used to obtain the inclination angle of the tangent line of the brake pad edge curve. The inclination angle of the tangent line of the brake pad edge curve is then fitted using the least squares method to obtain the distribution function f(x). ， Where x is the inclination angle of the tangent, θ i Here, n is the number of samples; P32. Based on the distribution function f(x), perform a two-dimensional discrete Fourier transform to obtain a periodic signal of the inclination angle of the tangent line of the brake pad edge curve; P33. Based on the periodic signal of the inclination angle of the tangent of the brake pad edge curve, the MTF value of the corresponding brake pad is output through normalized modulus. In step P2, the EIAJ algorithm includes: P21. Based on the first image data information of the brake pad and the second image data information of the brake pad, a large number of candidate regions are generated from the first and second image data information of the brake pad, these regions are filtered, and candidate regions containing brake pad images are output. P22. Based on the candidate regions containing the brake pad image, train a network to classify the candidate regions containing the brake pad image, extract image features, and output image feature information of different regions of the brake pad image; P23. Based on the image feature information of different regions of the brake pad, the feature data information of the edge region of the brake pad is obtained by calculating the gradient map.

2. The brake pad detection method based on a CCD camera according to claim 1, characterized in that, In step P5, the hole detection algorithm includes: P51. Preprocess the first image data information and the second image data information of the brake pad to obtain the processed first and second image data information of the brake pad; P52. Based on the processed brake pad first image data information, match the value box to obtain the value of each pixel in the value box of the brake pad first image as Y1. Based on the processed brake pad second image data information, match the value box to obtain the value of each pixel in the value box of the brake pad second image as Y2. P53. Based on the value Y1 of each pixel within the value frame of the first image of the brake pad and the value Y2 of each pixel within the value frame of the second image of the brake pad, according to the contrast function C ij , This gives the contrast of the current pixel.

3. The brake pad detection method based on a CCD camera according to claim 2, characterized in that: The preprocessing involves performing gamma correction on the image.

4. The brake pad detection method based on a CCD camera according to claim 2, characterized in that: The range of the first image frame of the brake pad is 1 pixel, 17*17 pixels, 25*25 pixels, and 35*35 pixels; the range of the second image frame of the brake pad corresponding to the first image frame of the brake pad is 31*31 pixels, 85*85 pixels, 125*125 pixels, and 175*175 pixels.

5. The brake pad detection method based on a CCD camera according to claim 2, characterized in that, The hole detection algorithm further includes: if it is a hole or crack, performing binarization processing, calculating the area of ​​the hole or crack, and then filtering and judging according to the judgment criteria.

6. A brake pad detection system based on a CCD camera, comprising computer equipment, characterized in that, The computer device is programmed or configured to perform the steps of the brake pad detection method based on a CCD camera as described in any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is programmed or configured to perform the brake pad detection method based on a CCD camera as described in any one of claims 1 to 5.

Images