A method for monitoring temperature and wear of a brake disc of a high-speed train

Abstract

The application discloses a kind of high-speed train brake disc temperature and abrasion amount monitoring method, including high-definition camera, laser emitter, control box, infrared thermal imager, DSP minimum system board, power supply step-down module, ZigBee router, by DSP minimum system board control laser emitter is emitted laser beam to brake disc, form light section curve, pass through high-definition camera and shoot light section curve, again transmission is given to DSP minimum system board and carries out image processing, by calculating the pixel number of brake disc to obtain the thickness of brake disc, compare the measured thickness with original thickness to calculate the abrasion amount of brake disc, by infrared thermal imager acquisition brake disc infrared image, and transmission is given to DSP minimum system board and handles, and then calculate brake surface temperature information, and by DSP minimum system board control ZigBee router will be calculated out brake disc wear, current temperature, defect condition Real-time transmission is sent to the PC of control console and carries out real-time display.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of brake disc temperature and wear amount monitoring, in particular to a high-speed train brake disc temperature and wear amount monitoring method. BACKGROUND

[0002] High-speed railway has the advantages of safety and comfort, energy saving and environmental protection, rapidness and convenience, large transportation capacity, etc., and plays a huge role in economic development and people's life, and is paid more and more attention by more and more countries. As one of the core technologies of high-speed train, the braking technology is directly related to the safety, stability and comfort of train operation. In the development process of high-speed train, the braking technology has become one of the core problems limiting the high-speed of train. Through continuous accumulation of own technical achievements and introduction and absorption of foreign advanced technology, great progress has been made in wheel-rail system and braking system, etc., which enables the rapid rise of high-speed railway in China.

[0003] The train braking system is realized through the friction between the brake disc and the brake pad. In the train braking process, the high friction heat of the high-speed braking interface causes the temperature of the friction surface to rise sharply, and the temperature change amplitude of the brake disc in the friction braking process of the brake pad can reach 400-500 DEG C. The continuous increase of train speed inevitably leads to the increase of kinetic energy, which causes the heat energy generated by the brake disc during braking to increase sharply, resulting in a very large thermal load on the brake disc, and the cyclic alternating thermal stress also causes fatigue cracks in the brake disc. In addition, the low temperature service environment in winter further increases the temperature change amplitude of the friction surface of the brake disc, which has a great influence on the friction and wear mechanism of the brake disc and brake pad, and makes it more likely to cause abnormal wear of the brake disc and brake pad in the low temperature environment, seriously affecting the stability and reliability of the braking system during service.

[0004] In the prior art, there are many monitoring and researches on the wear, defects and heating of wheelsets, but few are applied to high-speed train brake discs. The existing methods for detecting defects of brake discs mainly include ultrasonic detection method, magnetic powder flaw detection method and eddy current detection method. These methods need to be detected after stopping, and cannot realize detection in the running state of the vehicle. SUMMARY

[0005] In order to solve the above problems, the purpose of the present application is to provide a high-speed train brake disc temperature and wear amount monitoring method, which collects the infrared radiation energy emitted by the brake disc through an infrared thermal imager, forms an infrared image, and then processes the infrared thermal imaging picture to identify the temperature value on the brake disc. In addition, the image measurement technology and laser line light source are combined to detect the wear state of the brake disc.

[0006] In order to achieve the above-mentioned purpose, the application provides a kind of high-speed train brake disc temperature and wear monitoring method, which is realized as follows:

[0007] A kind of high-speed train brake disc temperature and wear monitoring method, including high-definition camera, laser emitter, control box, infrared thermal imager, DSP minimum system board is placed in control box, power supply step-down module, ZigBee router, high-definition camera, laser emitter, control box, infrared thermal imager are all installed in the car bottom of brake disc oblique upper side, and high-definition camera, laser emitter and brake disc are in the same horizontal plane, the included angle between infrared thermal imager and brake disc is 45 degrees, power supply step-down module is reduced to 12V after the power supply of high-speed train and is powered for high-definition camera, laser emitter, infrared thermal imager, ZigBee router, DSP minimum system board, image measurement technology and laser line light source are combined to monitor the wear state of brake disc, laser emitter is controlled by DSP minimum system board to brake disc and emits laser beam, forms light section curve, high-definition camera is photographed to light section curve, then is transmitted to DSP minimum system board and is handled, the thickness of brake disc is calculated by calculating the pixel number of brake disc, the wear of brake disc is calculated by comparing the measured thickness with original thickness, the infrared image of brake disc is collected by infrared thermal imager, and is transmitted to DSP minimum system board and is handled, and then the temperature information of brake surface is calculated, and ZigBee router is controlled by DSP minimum system board to send brake disc wear, current temperature and defect condition calculated in real time to PC machine of control console and display in real time.

[0008] The ZigBee router of the application transmits data to PC machine of control console and displays in real time, ZigBee coordinator is installed on train control console, for receiving information transmitted by ZigBee router, and transmitting information to PC machine, and displaying data in real time in PC machine.

[0009] The scheme for identifying brake disc temperature according to infrared thermal imaging picture is as follows:

[0010] S1. image preprocessing

[0011] ① read the infrared image to be identified, and perform gray scale transformation and Gamma correction;

[0012] ② draw the histogram of the infrared image to be identified, and adaptively obtain the threshold value on the right side of the histogram trough;

[0013] ③ adopt adaptive threshold to perform binaryzation processing on the image;

[0014] S2. image segmentation

[0015] ① The pixel accumulation method is used to locate the rectangular frame on the binary image;

[0016] ② The ROI region is determined according to the position information;

[0017] ③ The vertical projection method is used for character segmentation of the ROI region, and a temperature value data set is established.

[0018] S3. Temperature identification

[0019] ① A CNN network with a depth of 7 is built, and the network parameters are determined, and the temperature value training set and test set are divided according to the ratio of 8:2;

[0020] ② The CNN network is trained and tested, and the character recognition results are analyzed;

[0021] ③ The brake disc infrared image temperature value recognition and recording system is designed and selected, and a plurality of infrared images are tested for temperature identification results.

[0022] The scheme of the high-definition camera collecting the light section curve is that the laser emitter emits a laser beam to irradiate on the brake disc to form a light section curve, when the high-definition camera shoots the light section curve, the included angle between the center line of the high-definition camera and the light plane of the laser beam is θ, the object distance when the high-definition camera shoots is μ, the image distance is ν, the focal length is f, the actual thickness of the brake disc is L, the magnification is β, the pixel number is n, the unit pixel length is p, the image length is l, and k represents the object image proportionality coefficient, so that:

[0023]

[0024]

[0025] l=n·p (3)

[0026] The object image proportionality coefficient k can be obtained from formula (1) (2) (3):

[0027]

[0028] The actual thickness L of the brake disc is obtained as:

[0029]

[0030] After the actual light section curve is processed by the DSP minimum system board, the pixel number of the brake disc in the image is obtained, and the thickness of the brake disc can be calculated.

[0031] The scheme of the DSP minimum system board of the application for processing the light section curve image collected by the high-definition camera is:

[0032] S1. Image processing

[0033] (1) Image type conversion

[0034] The collected image is subjected to gray processing, and the collected color image is converted into a gray image to improve the subsequent processing speed.

[0035] (2) Image negative

[0036] In the collected image, the gray scale of the light band contour region is relatively bright, and the gray scale of other regions is relatively dark. In order to further facilitate subsequent image processing, the gray scale of the light band contour region is made dark and the gray scale of other regions is made bright by using the negative of the image. The negative of the image is realized by using a gray scale transformation equation, which is specifically:

[0037] D B =f(D A )=f(A)·D A +f(B) (6)

[0038] In the formula, f(A) represents the slope of the linear function, f(B) represents the intercept of the linear function on the Y axis, D A represents the gray scale of the input image, and D B represents the gray scale of the output image. When f(A) > 1, the contrast of the output image will increase, when f(A) < 1, the contrast of the output image will decrease, when f(A) = 1 and f(B) ≠ 0, the gray scale value of all pixels will increase or decrease, and the corresponding image will be darker or brighter, when f(A) < 0, the dark area will become lighter and the bright area will become darker, when f(A) = 1 and f(B) = 0, the input image and the output image are the same, when f(A) = -1 and f(B) = 255, the output image and the output image are exactly opposite in gray scale value, so f(A) in formula (6) takes the value -1 and f(B) takes the value 255;

[0039] (3) Image filtering and denoising

[0040] Since the collected brake disc image has certain noise interference, it will affect the accuracy of the image recognition result, so a two-dimensional median filter is used for image filtering, which is specifically:

[0041] 1) First, establish a filter window gray scale histogram of the first pixel point to be processed, find the median of all pixel gray scale values in the window, and calculate the number of gray scale values less than the median in the window.

[0042] 2) Move the window center to the next pixel point, update the gray scale histogram to the median of the pixel gray scale values in the current window, and update the number of gray scale values less than the median in the current window.

[0043] 3) According to the relationship between the number of gray scale values less than the median value in the current window and half of the total number of pixels in the window, the gray scale median value of the current window is found upwards or downwards;

[0044] 4) If the current pixel point is not the last one in a row, go to step 2); if the current pixel point is the last one in a row and is not the last row to be processed, move the window center to the pixel point in the first column of the next row, and then go to step 1); otherwise, the process ends.

[0045] S2. Image analysis

[0046] After image processing on the collected brake disc image, the brake disc contour curve can be obtained, and by comparison with the contour curve of the standard brake disc and through image size measurement, the actual brake disc thickness and wear amount can be measured, and the calculation formula of brake disc wear is:

[0047] B M =B-L B (7)

[0048] In the formula, L B is the actual measured thickness value of the brake disc, B represents the standard value of the brake disc thickness, and B M represents the wear value of the brake disc;

[0049] After collecting the contour map of the brake disc to be measured and the contour map of the standard brake disc, image processing is performed respectively, then the curves of the contour map of the brake disc to be measured and the contour map of the standard brake disc are scanned respectively to obtain the vertex of the brake disc rim, and the contour map of the brake disc to be measured and the contour map of the standard brake disc are overlapped by taking the vertex as a reference and using the Laplace fusion algorithm, and the difference between the y values of the two curves at the offset line L A of 70mm in the x direction is measured by taking the intersection 0 of the offset line L A and the standard surface as a reference, and the difference between the x values of the two curves at the offset line of 0mm in the positive y direction is measured, and then the wear amount of the brake disc is calculated.

[0050] Since the present application detects the temperature information of the brake disc by using an infrared thermal imager, and detects the wear state of the brake disc by using the combination of image measurement technology and laser line light source, the following beneficial effects can be obtained:

[0051] 1. The infrared thermal imager is used to detect the temperature information of the brake disc, and can realize non-contact measurement, and compared with contact measurement, the infrared thermal imager is not affected by the friction of the brake disc, thereby improving the service life.

[0052] 2. The image measurement technology and laser line light source are combined to detect the wear state of the brake disc, the wear of the brake disc can be monitored in real time, the real-time display can be realized on the PC end of the console, the wear amount, temperature information, defect condition and the like of the brake disc can be more directly viewed. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 It is an installation structure schematic view of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0054] Figure 2 It is a structure schematic view of the control box of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0055] Figure 3 It is a working principle view of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0056] Figure 4 It is a scheme flow chart of the infrared thermal imaging picture identification brake disc temperature of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0057] Figure 5 It is an imaging light path view of the high-definition camera collection light section curve of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0058] Figure 6 It is an image anti-color flow chart of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0059] Figure 7 It is an image filtering denoising flow chart of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0060] Figure 8 It is an image analysis scheme flow chart of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0061] Figure 9 It is a wear amount measurement mathematical model of the monitoring method for the temperature and wear amount of the high-speed train brake disc.

[0062] Main element symbol explanation.

[0063] High-definition camera 1 Laser emitter 2 Control box 3 Infrared thermal imager 4 DSP minimum system board 5 Power supply step-down module 6 ZigBee router 7 ZigBee coordinator 8 PC 9 DETAILED DESCRIPTION

[0064] The application will be further described in detail below in combination with the embodiments and the drawings.

[0065] Please refer to Figures 1 to 9The monitoring method of the brake disc temperature and wear of a high-speed train in the application comprises a high-definition camera 1, a laser emitter 2, a control box 3, and an infrared thermal imager 4.

[0066] As shown in Figure 1 , Figure 2 , the control box 3 is placed with a DSP minimum system board 5, a power supply step-down module 6, and a ZigBee router 7, the high-definition camera 1, the laser emitter 2, the control box 3, and the infrared thermal imager 4 are all installed on the bottom of the train above the brake disc, the high-definition camera 1 and the laser emitter 2 are on the same horizontal plane with the brake disc, the angle between the infrared thermal imager 4 and the brake disc is 45 degrees, the high-definition camera 1, the laser emitter 2, the infrared thermal imager 4, and the ZigBee router 7 are electrically connected with the DSP minimum system board 5, the power supply step-down module 6 steps down the power supply of the locomotive to 12V to supply power for the high-definition camera 1, the laser emitter 2, the infrared thermal imager 4, the ZigBee router 7, and the DSP minimum system board 5, the image measurement technology and the laser line light source are combined to detect the wear state of the brake disc, the laser emitter 2 is controlled by the DSP minimum system board 5 to emit laser beams to the brake disc to form a light section curve, the light section curve is photographed by the high-definition camera 1 and then transmitted to the DSP minimum system board 5 for image processing, the thickness of the brake disc is calculated by counting the pixel number of the brake disc, the wear amount of the brake disc is calculated by comparing the measured thickness with the original thickness, the infrared image of the brake disc is collected by the infrared thermal imager 4 and then transmitted to the DSP minimum system board 5 for processing, the temperature information of the brake surface is calculated, and the ZigBee router 7 is controlled by the DSP minimum system board 5 to transmit the calculated wear amount of the brake disc, the current temperature, and the defect condition to the PC machine 9 of the control console in real time for real-time display.

[0067] As shown in Figure 3 , the ZigBee router 7 transmits data to the PC machine 9 of the control console for real-time display, a ZigBee coordinator 8 is installed on the train control console to receive the information transmitted by the ZigBee router 7 and transmit the information to the PC machine 9, and the data is displayed in real time in the PC machine 9.

[0068] As shown in Figure 4 , the scheme for identifying the temperature of the brake disc according to the infrared thermal imaging picture of the application is as follows:

[0069] S1. Image preprocessing

[0070] ① Read the infrared image to be identified, perform gray scale conversion and Gamma correction;

[0071] ② Draw a histogram of the infrared image to be identified, and adaptively obtain a threshold value on the right side of the histogram valley;

[0072] ③Adopt adaptive threshold to binarize the image;

[0073] S2. Image segmentation

[0074] ①Adopt pixel accumulation method to locate the rectangular frame on the binarized image;

[0075] ②Determine the ROI region according to the position information;

[0076] ③Adopt vertical projection method to segment the characters in the ROI region, and establish the temperature value dataset.

[0077] S3. Temperature identification

[0078] ①Build a CNN network with a depth of 7, determine the network parameters, and divide the temperature value training set and test set according to the ratio of 8:2;

[0079] ②Train and test the CNN network, and analyze the character recognition results;

[0080] ③Combine the design of the brake disc infrared image temperature value identification and recording system, and select several infrared images to test the temperature identification results.

[0081] As shown in Figure 5 , the scheme of the high-definition camera 1 collecting the light section curve is: the laser emitter 2 emits a laser beam to irradiate on the brake disc to form a light section curve, when the high-definition camera 1 shoots the light section curve, the angle between the center line of the high-definition camera 1 and the light plane of the laser beam is θ, assuming that the object distance when the high-definition camera 1 shoots is μ, the image distance is v, the focal length is f, the actual thickness of the brake disc is L, the magnification is β, the pixel number is n, the unit pixel length is p, the image length is l, and k represents the object-image proportionality coefficient, then:

[0082]

[0083]

[0084] l=n·p (3)

[0085] From equations (1), (2) and (3), the object-image proportionality coefficient k is:

[0086]

[0087] The actual thickness L of the brake disc is:

[0088]

[0089] After the actual light section curve is processed by the DSP minimum system board 5, the pixel number of the brake disc in the image is obtained, and the thickness of the brake disc can be calculated.

[0090] The DSP minimum system board 5 processes the light section curve image collected by the high-definition camera 1 according to the following scheme:

[0091] S1. Image processing

[0092] (1) Image type conversion

[0093] The collected image is subjected to grayscale processing, and the collected color image is converted into a grayscale image to improve the subsequent processing speed.

[0094] (2) Image inversion

[0095] As shown in Figure 6 , the collected image, the gray scale of the light band profile area is relatively bright, and the gray scale of other areas is relatively dark. In order to further facilitate subsequent image processing, the gray scale of the light band profile area is made to be relatively dark and the gray scale of other areas is made to be relatively bright by using image inversion. The gray scale of the image is realized by using a gray scale transformation equation, which is specifically:

[0096] D B =f(D A )=f(A)·D A +f(B) (6)

[0097] In the formula, f(A) represents the slope of the linear function, f(B) represents the intercept of the linear function on the Y-axis, D A represents the gray scale of the input image, and D B represents the gray scale of the output image. When f(A)>1, the contrast of the output image will increase, when f(A)<1, the contrast of the output image will decrease, when f(A)=1 and f(B)≠0, the gray scale value of all pixels will increase or decrease, and the corresponding image will be darker or brighter, when f(A)<0, the dark area will become brighter and the bright area will become darker, when f(A)=1 and f(B)=0, the input image and the output image are the same, when f(A)=-1 and f(B)=255, the output image and the output image are exactly opposite in gray scale value, so f(A) in formula (6) is-1 and f(B) is 255;

[0098] (3) Image filtering and denoising

[0099] As shown in Figure 7 , since the collected brake disc image has certain noise interference, it will affect the accuracy of the image recognition result, so a two-dimensional median filter is used for image filtering, which is specifically:

[0100] 1) First, establish a filter window gray scale histogram of the first pixel point to be processed, find the median value of all pixel gray scale values in the window, and calculate the number of gray scale values less than the median value in the window;

[0101] 2) Move the center of the window to the next pixel, update the grayscale histogram to the median grayscale value of the pixels in the current window, and update the number of grayscale values ​​in the current window that are less than the median value.

[0102] 3) Based on the relationship between the number of gray values ​​smaller than the median value in the current window and half the total number of pixels in the window, search upwards or downwards for the median gray value of the current window;

[0103] 4) If the current pixel is not the last pixel in a row, proceed to step 2); if the current pixel is the last pixel in a row but not the last row to be processed, move the center of the window to the first pixel of the next row and then proceed to step 1); otherwise, the process ends.

[0104] S2. Image Analysis

[0105] like Figure 9 As shown, image processing of the acquired brake disc image yields the brake disc's outline curve. By comparing this curve with the outline curve of a standard brake disc and measuring the image dimensions, the actual brake disc thickness and wear can be determined. The formula for calculating brake disc wear is:

[0106] B M =BL B (7)

[0107] In the formula, L B For the actual measured thickness of the brake disc, B represents the standard value of the brake disc thickness. M This indicates the wear value of the brake disc;

[0108] like Figure 8 As shown, after acquiring the outline images of the brake disc under test and the standard brake disc, image processing was performed on each. Then, the curves of both the brake disc under test and the standard brake disc outline images were scanned to obtain the vertex of the brake disc rim. Using this vertex as a reference, the Laplacian fusion algorithm was used to overlap the brake disc under test and the standard brake disc outline images. Using the inner surface of the brake disc as a reference, the difference in y-value between the two curves was measured at a 70 mm offset in the x-direction. The difference in y-value between the two curves was then calculated using this offset line L. A Using the intersection point 0 with the standard surface as a reference, the difference in x-values ​​between the two curves at a positive y-direction offset of 0 mm is measured, and the wear of the brake disc is then calculated.

[0109] The working principle and process of this invention are as follows:

[0110] like Figure 3As shown, the laser emitter 2 is controlled by the DSP minimum system board 5 to emit a laser beam to the brake disc to form a light section curve, the light section curve is photographed by the high-definition camera 1 and then transmitted to the DSP minimum system board 5 for image processing, the thickness of the brake disc is calculated by counting the pixel number of the brake disc, the wear amount of the brake disc is calculated by comparing the measured thickness with the original thickness, the infrared image of the brake disc is collected by the infrared thermal imager 4 and then transmitted to the DSP minimum system board 5 for processing, the temperature information of the brake surface is calculated, the calculated wear amount of the brake disc, the current temperature and the defect condition are sent to the ZigBee coordinator 8 in real time by the ZigBee router 7 controlled by the DSP minimum system board 5, the information is transmitted to the PC 9 by the ZigBee coordinator 8, and the real-time display of data is realized in the PC 9.

Claims

1. A method for monitoring the temperature and wear of a high-speed train brake disc, characterized in that: The application is applied to a monitoring device, the monitoring device comprises a high-definition camera, a laser emitter, a control box, an infrared thermal imager, a DSP minimum system board, a power down module, a ZigBee router are placed in the control box, the high-definition camera, the laser emitter, the control box and the infrared thermal imager are all installed on the bottom of the vehicle above the brake disc, the high-definition camera and the laser emitter are on the same horizontal plane with the brake disc, the angle between the infrared thermal imager and the brake disc is 45 degrees, the power down module reduces the power of the locomotive to 12V to supply power for the high-definition camera, the laser emitter, the infrared thermal imager, the ZigBee router and the DSP minimum system board, the wear state of the brake disc is detected by combining the image measurement technology and the laser line light source, the laser emitter is controlled by the DSP minimum system board to emit a laser beam to the brake disc to form a light section curve, the high-definition camera photographs the light section curve and then transmits the image to the DSP minimum system board for image processing, the thickness of the brake disc is calculated by counting the pixel number of the brake disc, the wear amount of the brake disc is calculated by comparing the measured thickness with the original thickness, the infrared image of the brake disc is collected by the infrared thermal imager and then transmitted to the DSP minimum system board for processing, the temperature information of the brake surface is calculated, and the ZigBee router is controlled by the DSP minimum system board to send the calculated wear amount of the brake disc, the current temperature and the defect condition to the PC of the control console for real-time display; The scheme for collecting the infrared image of the brake disc by the infrared thermal imager and transmitting the image to the DSP minimum system board for processing to calculate the temperature information of the brake surface is as follows: S1. image preprocessing ① read the infrared image to be recognized, perform gray scale conversion and Gamma correction; ② draw a histogram of the infrared image to be recognized, and adaptively obtain a threshold value on the right side of the histogram valley; ③ perform binaryzation processing on the image by using the adaptive threshold value; S2. image segmentation ① locate a rectangular frame on the binaryzation image by using the pixel accumulation method; ② determine the ROI region according to the position information; ③ perform character segmentation on the ROI region by using the vertical projection method, and establish a temperature value data set; S3. temperature recognition ① build a CNN network with a depth of 7, determine the network parameters, and divide the temperature value training set and the test set according to the ratio of 8:2; ② train and test the CNN network, and analyze the character recognition results; ③ combine the design of the brake disc infrared image temperature value recognition and recording system, and select a plurality of infrared images to test the temperature recognition results; The scheme of the high-definition camera collecting the light section curve is that a laser emitter emits a laser beam to irradiate on the brake disc to form a light section curve, and when the high-definition camera shoots the light section curve, the angle between the center line of the high-definition camera and the light plane of the laser beam is , the object distance when the high-definition camera shoots is , the image distance is , the focal length is , the actual thickness of the brake disc is , the magnification is , the pixel number is , the unit pixel length is , the image length is , represents the object image proportionality coefficient, and then: （1） （2） （3） From equations (1), (2), (3) the object-image scale factor is derived is: （4） Deriving the actual thickness of the brake disc is: （5） After the actual light section curve is processed by the DSP minimum system board, the pixel number of the brake disc in the image is obtained, and the thickness of the brake disc is calculated.

2. The method for monitoring the temperature and wear of high-speed train brake discs according to claim 1, characterized in that: The scheme for processing the light section curve image collected by the high-definition camera by the DSP minimum system board is as follows: S1. image processing (1) image type conversion perform gray scale processing on the collected image, and convert the collected color image into a gray scale image to improve the subsequent processing speed; (2) image inversion In the collected image, the gray scale of the light band profile area is relatively bright, while the gray scale of other areas is relatively dark. In order to further facilitate subsequent image processing, the gray scale of the light band profile area is made to be dark and the gray scale of other areas is made to be bright by using the negative color of the image. The negative color of the image is realized by using a gray scale transformation equation, specifically: （6） wherein, represents the slope of the linear function, represents the intercept of the linear function on the Y-axis, represents the gray level of the input image, represents the gray level of the output image, when the contrast of the output image will increase, when the contrast of the output image will decrease, when and the gray level of all pixels will increase or decrease, and the corresponding image will be darker or brighter, when the dark areas will become lighter and the light areas darker, when and the input image and the output image are identical, when and the output image is the inverse of the output image in terms of gray level values, so that takes the value -1 in equation (6), takes the value 255; (3) Image filtering and denoising Since the collected brake disc image has certain noise interference, which will affect the accuracy of the image recognition result, a two-dimensional median filter is used for image filtering, specifically: 1) First, establish the filter window gray scale histogram of the first pixel point to be processed, find the median of all pixel gray values in the window, and calculate the number of gray values less than the median in the window; 2) Move the window center to the next pixel point, update the gray scale histogram to the median of the pixel gray values in the current window, and update the number of gray values less than the median in the current window; 3) According to the relationship between the number of gray values less than the median in the current window and half of the total number of window pixel points, find the gray median of the current window upward or downward; 4) If the current pixel point is not the last one in a row, go to step 2); if the current pixel point is the last one in a row and is not the last row to be processed, move the window center to the pixel point in the first column of the next row, and go to step 1); otherwise, the process is ended; S2. Image analysis After image processing of the collected brake disc image, the brake disc contour curve can be obtained. By comparing with the standard brake disc contour curve and through image size measurement, the actual brake disc thickness and wear can be measured. The calculation formula of brake disc wear is: （7） wherein is the actual measured thickness value of the brake disc, is the standard value of the thickness of the brake disc, is the wear value of the brake disc; After collecting the profile graph of the brake disc to be tested and the profile graph of the standard brake disc, image processing is performed respectively, then the curves of the profile graph of the brake disc to be tested and the profile graph of the standard brake disc are scanned respectively to obtain the vertex of the brake disc rim, taking the vertex as the reference, the profile graph of the brake disc to be tested and the profile graph of the standard brake disc are overlapped by using the Laplace fusion algorithm, taking the inner side surface of the brake disc as the reference, the difference between the y values of the two curves at the place 70 mm away from the x direction is measured, taking the intersection 0 of the standard surface as the reference, the difference between the x values of the two curves at the place 0 mm away from the positive y direction is measured, and then the wear amount of the brake disc is calculated. The wear amount of the brake disc is calculated.

Images