Automobile paint wear resistance detection device and method

Abstract

The invention discloses an automobile paint wear resistance detection device and method, and relates to the technical field of automobile paint wear resistance detection, and the automobile paint wear resistance detection device comprises an automobile paint surface fixing tool, a paint surface friction simulation device, a surrounding type scratch visual detection device and a wear loss analysis unit based on a computer. The paint surface friction simulation device performs physical friction on a target automobile paint surface according to set friction parameters, simulates different physical friction possibly existing in the automobile paint using process, and collects a plurality of visual pictures before and after physical friction of the target automobile paint surface through the surrounding type scratch visual detection device; the abrasion loss analysis unit based on the computer adopts a deep learning model to analyze the abrasion loss of each visual picture, the maximum value and the average value of all the abrasion loss are compared and analyzed with a standard abrasion loss threshold value, and the abrasion resistance detection score of the target automobile paint surface is output, so that the abrasion resistance of the automobile paint is quantitatively and automatically detected; the detection efficiency is high.

Description

Technical Field

[0001] The invention relates to the technical field of automobile paint wear resistance detection, and in particular to an automobile paint wear resistance detection device and method. Background Art

[0002] At present, the detection of automobile paint defects mainly relies on manual inspection, which is costly and subjective, limiting the improvement of production efficiency.

[0003] With the development of industrial informatization and intelligence, paint defect detection technology based on machine vision has gradually gained attention. This technology simulates the visual function of the human eye and can assist in the detection and judgment of paint defects, thus improving detection efficiency and accuracy.

[0004] Manual testing has the problems of high subjectivity, low efficiency and high cost.

[0005] Although automated detection technology has developed, its recognition rate and adaptability still need to be improved.

[0006] A Chinese invention patent application with publication number CN117990546A discloses a paint wear resistance detection device, comprising a body, a paint cake placed in the body, a detection device comprising a turntable, a detection component, a second rotation drive device and a second lifting mechanism, the turntable is rotatably arranged on the body, the detection component is movably arranged on the turntable, and a grinding roller and a scraping head are provided on the detection component; the grinding roller is controlled to move and abut against the paint cake, and then the second rotation drive device is controlled to drive the turntable to rotate, and the first rotation drive device is controlled to drive the grinding roller to roll on the paint cake, so that the wear resistance of the paint under rolling friction can be measured; the first lifting mechanism is controlled to extend so that the scraping head can be driven to extend and abut against the paint cake, and then the turntable is controlled to rotate to drive the scraping head to rub on the paint cake, so that the wear resistance of the paint under sliding friction can be measured, thereby realizing the wear resistance detection of rolling friction and sliding friction of the paint.

[0007] The above-mentioned testing equipment can achieve the timed friction of the paint cake at the mechanical level, and then measure the amount of the paint cake rubbed off during this time, and the wear resistance of the paint under rolling friction can be obtained;

[0008] Finally, in order to express what method is used to detect the wear resistance of paint, it is urgent to design an automobile paint wear resistance detection device and method to realize automated and high-precision paint wear resistance detection. Summary of the invention

[0009] In order to solve the above technical problems, the present invention provides a wear resistance detection device and method for automobile paint. The following technical solutions are adopted:

[0010] A wear resistance detection device for automobile paint comprises an automobile paint surface fixing tool, a paint surface friction simulation device, a surround-type scratch visual detection device and a computer-based wear analysis unit. The target automobile paint surface can be detachably mounted on the top card slot of the automobile paint surface fixing tool. The paint surface friction simulation device comprises a plane moving component and a friction simulation component. The plane moving component is mounted on one side of the automobile paint surface fixing tool. The friction simulation component is mounted on the movable part of the plane moving component and is used to perform physical friction on the target automobile paint surface under set friction parameters. The surround-type scratch visual detection device is mounted above the target automobile paint surface and collects multiple visual images of the target automobile paint surface before and after physical friction at a set angle. The wear analysis unit communicates with the surround-type scratch visual detection device to interact with the visual images. The wear of each visual image is analyzed based on a deep learning model. The maximum value and average value of all wear amounts are taken and compared with a standard wear amount threshold value for analysis, and the wear resistance detection score of the target automobile paint surface is output.

[0011] By adopting the above technical solution, a car paint surface fixing tool is used to install the target car paint surface. The target car paint surface can be in the form of a standard sheet metal with a corresponding paint surface to be tested, for example, a sheet metal surface of 200mm by 200mm is sprayed with 0 microns of the car paint to be tested according to the standard car paint. The car paint surface fixing tool can be in the form of a metal part fixed on a workbench, and a clamping groove is provided on the top surface. The standard sheet metal of the target car paint surface can be clamped in the clamping groove of the car paint surface fixing tool. The paint surface friction simulation device can perform physical friction on the target car paint surface according to the set friction parameters to simulate different physical frictions that may exist during the use of the car paint. Based on the wear generated thereby, a plurality of visual images of the target car paint surface before and after physical friction are collected by a surround scratch visual detection device. Finally, a computer-based wear analysis unit uses a deep learning model to analyze the wear of each visual image, and takes the maximum value and average value of all wear amounts and compares and analyzes with the standard wear amount threshold, and outputs the wear resistance detection score of the target car paint surface, thereby quantitatively and automatically detecting the wear resistance of the car paint surface, and the detection efficiency is high.

[0012] Optionally, the paint surface friction simulation device includes a support member, a plane moving component and a friction simulation component, the support member is installed on the side of the automobile paint surface fixing tooling, the plane moving component includes a longitudinal electric moving mechanism and a transverse electric moving mechanism, the track of the longitudinal electric moving mechanism is installed on the top surface of the support member, and the track of the transverse electric moving mechanism is installed on the slider of the longitudinal electric moving mechanism, the friction simulation component includes a friction simulation bracket, a servo electric cylinder, a friction application block and a detachable friction plate assembly, the bottom of the friction simulation bracket is installed on the slider of the transverse electric moving mechanism, the base of the servo electric cylinder is installed at the end of the friction simulation bracket with the piston rod facing vertically downward, the top surface of the friction application block is installed on the end of the piston rod of the servo electric cylinder, and the detachable friction plate assembly is detachably installed on the bottom surface of the friction application block.

[0013] Optionally, the friction simulation bracket is an inverted L-shaped bracket.

[0014] Optionally, both the longitudinal electric moving mechanism and the transverse electric moving mechanism are electric servo linear guide rails.

[0015] By adopting the above technical solution, the plane moving assembly is realized by two independent electric servo linear slide rails, and the longitudinal electric moving mechanism is mainly used to move the entire friction simulation assembly in and out of the top of the automobile paint surface fixing tooling, so as to facilitate the assembly of the automobile paint surface fixing tooling and the non-interference surrounding scratch visual detection device to capture the visual picture;

[0016] The lateral electric moving mechanism is mainly used to provide friction power. It can perform reciprocating linear motion at a set speed. After one linear motion reaches one end, it moves in the opposite direction at the same speed to the other end. The detachable friction plate assembly is used to simulate physical friction, and the servo cylinder can accurately apply pressure.

[0017] Optionally, the detachable friction plate assembly includes a friction plate mounting block and a friction plate, wherein the top surface of the friction plate mounting block is mounted on the end of the piston rod of the servo electric cylinder, and the friction plate is mounted on the bottom surface of the friction plate mounting block. When the friction plate contacts the paint surface of the target car, physical friction is simulated on the paint surface of the target car.

[0018] By adopting the above technical solution, the friction plate can be flexibly replaced according to the detection, for example, it can be made of rubber grinding wheel, wear-resistant paper, and steel wool.

[0019] Optionally, it also includes a friction simulation controller, which includes a touch screen, a buffer, a control chip and a servo driver. The touch screen is used to set friction parameters, and the friction parameters include pressure parameters, friction speed parameters and friction time parameters. The buffer is communicatively connected to the touch screen. The control chip communicates with the buffer to exchange pressure parameters, friction speed parameters and friction time parameters. Based on the pressure parameters, control instructions for controlling the servo electric cylinder are generated. Based on the friction speed parameters, control instructions for controlling the lateral electric moving mechanism are generated. The control chip controls the execution actions of the longitudinal electric moving mechanism, the lateral electric moving mechanism and the servo electric cylinder respectively through the servo driver.

[0020] By adopting the above technical solution, the touch screen can realize the rapid entry of pressure parameters, friction speed parameters and friction time parameters. These parameter control chips can be directly converted into control instructions for controlling the lateral electric moving mechanism and the servo electric cylinder respectively. The control instructions for the longitudinal electric moving mechanism can be implemented based on the position sensor.

[0021] Optionally, the surround-type visual detection device for scratches includes a visual bracket, an arc-shaped electric slide, a visual camera and a fill light, wherein the visual bracket is installed on the side of a fixed tooling on the paint surface of the automobile, the track of the arc-shaped electric slide is installed on the visual bracket, and the visual camera and the fill light are installed on a slider of the arc-shaped electric slide, and the control chip controls the arc-shaped electric slide, the visual camera and the fill light to perform actions respectively.

[0022] By adopting the above technical solution, the arc-shaped electric slide is installed on the visual bracket, which is actually located directly above the automobile paint surface fixing tooling. The lowest point of the arc-shaped electric slide is higher than the top of the friction simulation bracket of the paint surface friction simulation device, avoiding interference with the paint surface friction simulation device. When the longitudinal electric moving mechanism of the paint surface friction simulation device is actuated to move the paint surface friction simulation device out from above the target automobile paint surface, visual shooting can be carried out, and visual picture shooting at a set angle can be achieved, which can more comprehensively evaluate the wear resistance of the target automobile paint surface after physical friction simulation.

[0023] Optionally, the wear analysis unit includes a memory and a computer, wherein the memory communicates with the visual camera to exchange visual image data, and the computer analyzes the wear of each visual image based on a deep learning model and outputs a wear resistance detection score of the target car paint surface.

[0024] A method for detecting the wear resistance of automobile paint, using an automobile paint wear resistance detection device to perform physical friction simulation on a target automobile paint surface, and performing visual detection analysis to output a wear resistance detection score, comprising the following steps:

[0025] Step 1, select the type of friction plate and install the friction plate on the friction plate mounting block;

[0026] Step 2, setting the pressure parameter, friction speed parameter and friction time parameter through the touch screen;

[0027] Step 3, the friction simulation controller controls the longitudinal electric moving mechanism and the transverse electric moving mechanism to be linked, so that the detachable friction plate assembly is located at the center of the target car paint surface, controls the servo electric cylinder to apply a set pressure, and controls the transverse electric moving mechanism to perform a linear reciprocating motion at a set friction speed and for a set friction time;

[0028] Step 4: After step 3 is completed, the friction simulation controller controls the longitudinal electric moving mechanism to move the detachable friction plate assembly away from the target car paint surface;

[0029] Step 5: The friction simulation controller controls the slider movement of the arc-shaped electric slide from the leftmost to the rightmost, stopping for one to two seconds every 20° arc, and the camera and fill light are turned on to shoot the visual picture of the current angle. After shooting the visual pictures of all angles, all the visual pictures of all angles are packaged and sent to the wear analysis unit;

[0030] Step 6: The wear analysis unit analyzes the wear of each visual image based on the deep learning model and outputs the wear resistance detection score of the target car's paint surface.

[0031] Optionally, in step 6, a convolutional neural network is used to identify and classify different types of wear; and the degree of wear is quantified by analyzing image features, and the formula is: Where W is the wear amount, PX a is the number of pixels in the worn area identified by the deep learning model, PX all is the total number of pixels;

[0032] The formula for the wear test score is:

[0033] Where WL is the wear resistance test score, W max is the maximum wear of all visual images, W av is the mean wear value of all visual images, S is the wear threshold, α and β are weight coefficients, α is 0.6, and β is 0.4.

[0034] In summary, the present invention includes at least one of the following beneficial technical effects:

[0035] The present invention can provide a device and method for detecting the wear resistance of automobile paint. The standard sheet metal of the target automobile paint surface is clamped in the clamping groove of the automobile paint surface fixing tooling. The paint surface friction simulation device performs physical friction on the target automobile paint surface according to the set friction parameters to simulate the different physical frictions that may exist in the use process of the automobile paint. A surround scratch visual detection device is used to collect multiple visual images of the target automobile paint surface before and after the physical friction. The computer-based wear analysis unit uses a deep learning model to analyze the wear of each visual image, and takes the maximum value and average value of all wear amounts for comparison and analysis with the standard wear threshold, and outputs the wear resistance detection score of the target automobile paint surface, thereby quantitatively and automatically detecting the wear resistance of the automobile paint, and the detection efficiency is high. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the lateral structure principle of a wear-resistant detection device for automobile paint of the present invention;

[0037] Figure 2 This is a schematic diagram of the structural principle of a wear-resistant detection device for automobile paint according to the present invention;

[0038] Figure 3 The invention discloses a schematic diagram of the connection principle of electrical components of a wear-resistant detection device for automobile paint.

[0039] Explanation of the reference numerals: 11. Support member; 12. Longitudinal electric moving mechanism; 13. Transverse electric moving mechanism; 14. Friction simulation bracket; 15. Servo electric cylinder; 16. Friction application block; 17. Removable friction plate assembly; 21. Touch screen; 22. Buffer; 23. Control chip; 24. Servo drive; 31. Vision bracket; 32. Arc electric slide; 33. Vision camera; 34. Fill light; 41. Memory; 42. Computer; 100. Target car paint surface; 101. Car paint surface fixing tool. DETAILED DESCRIPTION

[0040] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0041] The embodiment of the invention discloses a device and method for detecting the wear resistance of automobile paint.

[0042] Reference Figure 1 - Figure 3Embodiment 1: A wear-resistance detection device for automobile paint comprises an automobile paint surface fixing tool 101, a paint surface friction simulation device, a surround-type scratch visual detection device and a computer-based wear analysis unit. The target automobile paint surface 100 can be detachably mounted on the top card slot of the automobile paint surface fixing tool 101. The paint surface friction simulation device comprises a plane moving component and a friction simulation component. The plane moving component is mounted on one side of the automobile paint surface fixing tool 101. The friction simulation component is mounted on the movable part of the plane moving component, and is used to perform physical friction on the target automobile paint surface 100 under set friction parameters. The surround-type scratch visual detection device is mounted above the target automobile paint surface 100, and multiple visual images of the target automobile paint surface 100 before and after physical friction are collected at a set angle. The wear analysis unit communicates with the surround-type scratch visual detection device to interact with the visual images, analyzes the wear of each visual image based on a deep learning model, takes the maximum value and average value of all wear amounts and compares and analyzes them with the standard wear amount threshold, and outputs the wear resistance detection score of the target automobile paint surface 100.

[0043] The target automobile paint surface 100 is installed by using the automobile paint surface fixing fixture 101. The target automobile paint surface 100 can be in the form of a standard sheet metal coated with a corresponding paint surface to be tested. For example, a sheet metal surface of 200 mm by 200 mm is sprayed with 180 microns of the paint to be tested according to the standard car paint. The automobile paint surface fixing fixture 101 can be in the form of a metal part fixed on a workbench, and a clamping groove is provided on the top surface. The standard sheet metal of the target automobile paint surface 100 can be clamped in the clamping groove of the automobile paint surface fixing fixture 101. The paint surface friction simulation device can be used to simulate the target automobile paint surface 100. 00Physical friction is performed according to the set friction parameters to simulate the different physical frictions that may exist in the use process of car paint. Based on the wear generated thereby, a plurality of visual images of the target car paint surface 100 before and after the physical friction are collected through a surround-type scratch visual detection device. Finally, a computer-based wear analysis unit uses a deep learning model to analyze the wear of each visual image, and takes the maximum and average values ​​of all wear amounts for comparative analysis with the standard wear threshold, and outputs the wear resistance detection score of the target car paint surface 100, thereby quantifying and automatically detecting the wear resistance of the car paint surface with high detection efficiency.

[0044] The paint surface friction simulation device includes a support member 11, a plane moving component and a friction simulation component. The support member 11 is installed on the side of the automobile paint surface fixing tool 101. The plane moving component includes a longitudinal electric moving mechanism 12 and a transverse electric moving mechanism 13. The track of the longitudinal electric moving mechanism 12 is installed on the top surface of the support member 11, and the track of the transverse electric moving mechanism 13 is installed on the slider of the longitudinal electric moving mechanism 12. The friction simulation component includes a friction simulation bracket 14, a servo electric cylinder 15, a friction applying block 16 and a detachable friction plate assembly 17. The bottom of the friction simulation bracket 14 is installed on the slider of the transverse electric moving mechanism 13, the base of the servo electric cylinder 15 is installed at the end of the friction simulation bracket 14, and the piston rod is vertically downward. The top surface of the friction applying block 16 is installed at the end of the piston rod of the servo electric cylinder 15, and the detachable friction plate assembly 17 is detachably installed on the bottom surface of the friction applying block 16.

[0045] The friction simulation bracket 14 is an inverted L-shaped bracket.

[0046] The longitudinal electric moving mechanism 12 and the transverse electric moving mechanism 13 are both electric servo linear guide rails.

[0047] The plane moving assembly is realized by two independent electric servo linear slide rails, and the longitudinal electric moving mechanism 12 is mainly used to move the entire friction simulation assembly in and out of the automobile paint surface fixing fixture 101, so as to facilitate the assembly of the automobile paint surface fixing fixture 101 and the non-interference surrounding scratch visual detection device to capture the visual picture;

[0048] The lateral electric moving mechanism 13 is mainly used to provide friction power, and can perform reciprocating linear motion at a set speed. After one linear motion reaches one end, it moves in the opposite direction at the same speed to the other end. Through the detachable friction plate assembly 17, physical friction simulation is performed, and the servo cylinder 15 can accurately apply pressure.

[0049] The detachable friction plate assembly 17 includes a friction plate mounting block and a friction plate. The top surface of the friction plate mounting block is mounted on the end of the piston rod of the servo electric cylinder 15, and the friction plate is mounted on the bottom surface of the friction plate mounting block. When the friction plate contacts the surface of the target car paint surface 100, physical friction is simulated on the surface of the target car paint surface 100.

[0050] The friction plate can be flexibly replaced according to the test, for example, it can be made of rubber grinding wheel, wear-resistant paper, and steel wool.

[0051] It also includes a friction simulation controller, which includes a touch screen 21, a buffer 22, a control chip 23 and a servo driver 24. The touch screen 21 is used to set friction parameters, and the friction parameters include pressure parameters, friction speed parameters and friction time parameters. The buffer 22 is communicated with the touch screen 21, and the control chip 23 communicates with the buffer 22 to exchange pressure parameters, friction speed parameters and friction time parameters. Based on the pressure parameters, a control instruction for controlling the servo electric cylinder 15 is generated, and based on the friction speed parameters, a control instruction for controlling the transverse electric moving mechanism 13 is generated. The control chip 23 controls the execution actions of the longitudinal electric moving mechanism 12, the transverse electric moving mechanism 13 and the servo electric cylinder 15 respectively through the servo driver 24.

[0052] The touch screen 21 can realize the rapid entry of pressure parameters, friction speed parameters and friction time parameters. These parameter control chips 23 can directly convert them into control instructions for controlling the lateral electric moving mechanism 13 and the servo cylinder 15 respectively. The control instructions for the longitudinal electric moving mechanism 12 can be implemented based on the position sensor.

[0053] The surround-type scratch visual detection device includes a visual bracket 31, an arc-shaped electric slide 32, a visual camera 33 and a fill light 34. The visual bracket 31 is installed on the side of the automobile paint surface fixing tool 101, the track of the arc-shaped electric slide 32 is installed on the visual bracket 31, the visual camera 33 and the fill light 34 are installed on the slider of the arc-shaped electric slide 32, and the control chip 23 controls the arc-shaped electric slide 32, the visual camera 33 and the fill light 34 to perform actions respectively.

[0054] The arc-shaped electric slide 32 is installed on the visual bracket 31, which is actually located directly above the automobile paint surface fixing tool 101. The lowest point of the arc-shaped electric slide 32 is higher than the top of the friction simulation bracket 14 of the paint surface friction simulation device, avoiding interference with the paint surface friction simulation device. When the longitudinal electric moving mechanism 12 of the paint surface friction simulation device is actuated to move the paint surface friction simulation device out from above the target automobile paint surface 100, visual shooting can be performed, and visual picture shooting at a set angle can be achieved, which can more comprehensively evaluate the wear resistance of the target automobile paint surface 100 after physical friction simulation.

[0055] The wear analysis unit includes a memory 41 and a computer 42. The memory 41 communicates with the visual camera 33 to exchange visual image data. The computer 42 analyzes the wear of each visual image based on a deep learning model and outputs a wear resistance detection score of the target car paint surface 100.

[0056] A method for detecting wear resistance of automobile paint is provided, wherein a device for detecting wear resistance of automobile paint is used to perform physical friction simulation on a target automobile paint surface 100, and visual detection and analysis are performed to output a wear resistance detection score, and the method comprises the following steps:

[0057] Step 1, select the type of friction plate and install the friction plate on the friction plate mounting block;

[0058] Step 2, setting the pressure parameter, the friction speed parameter and the friction time parameter through the touch screen 21;

[0059] Step 3, the friction simulation controller controls the longitudinal electric moving mechanism 12 and the transverse electric moving mechanism 13 to move in a linkage manner, so that the detachable friction plate assembly 17 is located at the center of the target automobile paint surface 100, controls the servo electric cylinder 15 to apply a set pressure, and controls the transverse electric moving mechanism 13 to perform a linear reciprocating motion for a set friction time at a friction speed;

[0060] Step 4: After step 3 is completed, the friction simulation controller controls the longitudinal electric moving mechanism 12 to move the detachable friction plate assembly 17 away from the target automobile paint surface 100;

[0061] Step 5, the friction simulation controller controls the slider movement of the arc-shaped electric slide 32, from the leftmost to the rightmost, and stops for one to two seconds every 20° arc, and the camera 33 and the fill light 34 are turned on to shoot the visual picture of the current angle. After shooting the visual pictures of all angles, all the visual pictures of all angles are packaged and sent to the wear analysis unit;

[0062] Step 6: The wear analysis unit analyzes the wear of each visual image based on the deep learning model, and outputs the wear resistance detection score of the target car paint surface 100.

[0063] In step 6, a convolutional neural network is used to identify and classify different types of wear; and the degree of wear is quantified by analyzing image features. The formula is: Where W is the wear amount, PX a is the number of pixels in the worn area identified by the deep learning model, PX all is the total number of pixels;

[0064] The formula for the wear test score is:

[0065] Where WL is the wear resistance test score, W max is the maximum wear of all visual images, W av is the mean wear value of all visual images, S is the wear threshold, α and β are weight coefficients, α is 0.6, and β is 0.4.

[0066] The process of identifying the number of pixels in the worn area using deep learning models, especially convolutional neural networks (CNNs), is divided into the following steps:

[0067] Resize the image to a uniform size and normalize it;

[0068] Increase data diversity and improve the generalization ability of the model through operations such as rotation, scaling, and flipping;

[0069] Use multiple convolutional layers to extract features from the image;

[0070] Reduce the size of the feature map through pooling operations while retaining important information;

[0071] Apply ReLU or other activation functions to increase the nonlinearity of the model;

[0072] In the last few layers of the network, fully connected layers are usually used to map features to corresponding categories.

[0073] Choose an appropriate loss function, such as binary cross entropy loss (for binary classification problems) or dice loss (for image segmentation problems);

[0074] Choose an optimizer, such as Adam or SGD, to update the weights of the network;

[0075] Use the dataset with annotations of worn areas to train a semantic segmentation network, such as U-Net, which is able to assign a class label to each pixel in the image.

[0076] Use the trained model to make predictions on the test image and output a segmentation map where each pixel is labeled as worn or non-worn.

[0077] Analyze the output segmentation map of the model and count the number of pixels marked as worn.

[0078] The calculation formula for the number of worn pixels is:

[0079]

[0080] Among them, PX a is the number of pixels in the area where wear occurs, H and W are the number of rows and columns of the segmentation map, respectively, and M(i, j) is the category label of the pixel at position (i, j) in the segmentation map.

[0081] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A wear-resistance detection device for automobile paint, characterized in that: The invention comprises a car paint surface fixing tool (101), a paint surface friction simulation device, a surround-type scratch visual detection device and a computer-based wear analysis unit. The target car paint surface (100) can be detachably mounted on the top card slot of the car paint surface fixing tool (101). The paint surface friction simulation device comprises a plane moving component and a friction simulation component. The plane moving component is mounted on one side of the car paint surface fixing tool (101). The friction simulation component is mounted on the movable part of the plane moving component and is used to perform physical friction on the target car paint surface (100) under set friction parameters. The surround-type scratch visual detection device is mounted above the target car paint surface (100) and collects a plurality of visual images of the target car paint surface (100) before and after physical friction at a set angle. The wear analysis unit communicates with the surround-type scratch visual detection device to exchange visual images. The wear of each visual image is analyzed based on a deep learning model. The maximum value and average value of all wear amounts are compared with a standard wear amount threshold value for analysis, and the wear resistance detection score of the target car paint surface (100) is output.

2. The automobile paint wear resistance detection device according to claim 1, characterized in that: The paint surface friction simulation device comprises a support member (11), a plane moving component and a friction simulation component, wherein the support member (11) is installed on the side of a vehicle paint surface fixing tool (101), the plane moving component comprises a longitudinal electric moving mechanism (12) and a transverse electric moving mechanism (13), the track of the longitudinal electric moving mechanism (12) is installed on the top surface of the support member (11), the track of the transverse electric moving mechanism (13) is installed on the slider of the longitudinal electric moving mechanism (12), and the friction simulation component comprises a friction simulation A bracket (14), a servo electric cylinder (15), a friction applying block (16) and a detachable friction plate assembly (17), wherein the bottom of the friction simulation bracket (14) is mounted on a slider of a lateral electric moving mechanism (13), the base of the servo electric cylinder (15) is mounted on the end of the friction simulation bracket (14), and the piston rod is vertically downward, the top surface of the friction applying block (16) is mounted on the end of the piston rod of the servo electric cylinder (15), and the detachable friction plate assembly (17) is detachably mounted on the bottom surface of the friction applying block (16).

3. The automobile paint wear resistance detection device according to claim 2, characterized in that: The longitudinal electric moving mechanism (12) and the transverse electric moving mechanism (13) are both electric servo linear slide rails.

4. The automobile paint wear resistance detection device according to claim 3, characterized in that: The friction simulation bracket (14) is an inverted L-shaped bracket.

5. The automobile paint wear resistance detection device according to claim 3, characterized in that: The detachable friction plate assembly (17) comprises a friction plate mounting block and a friction plate, wherein the top surface of the friction plate mounting block is mounted on the end of the piston rod of the servo electric cylinder (15), and the friction plate is mounted on the bottom surface of the friction plate mounting block. When the friction plate contacts the surface of the target automobile paint surface (100), physical friction is simulated on the surface of the target automobile paint surface (100).

6. The automobile paint wear resistance detection device according to claim 5, characterized in that: The invention also comprises a friction simulation controller, which comprises a touch screen (21), a buffer (22), a control chip (23) and a servo driver (24). The touch screen (21) is used to set friction parameters, which include pressure parameters, friction speed parameters and friction time parameters. The buffer (22) is connected to the touch screen (21) for communication. The control chip (23) communicates with the buffer (22) to exchange pressure parameters, friction speed parameters and friction time parameters. A control instruction for controlling the servo electric cylinder (15) is generated based on the pressure parameter. A control instruction for controlling the transverse electric moving mechanism (13) is generated based on the friction speed parameter. The control chip (23) controls the execution actions of the longitudinal electric moving mechanism (12), the transverse electric moving mechanism (13) and the servo electric cylinder (15) respectively through the servo driver (24).

7. The automobile paint wear resistance detection device according to claim 6, characterized in that: The surround type scratch visual detection device comprises a visual support (31), an arc-shaped electric slide (32), a visual camera (33) and a fill light (34), wherein the visual support (31) is mounted on the side of a fixing tool (101) for a paint surface of an automobile, the track of the arc-shaped electric slide (32) is mounted on the visual support (31), the visual camera (33) and the fill light (34) are mounted on a slider of the arc-shaped electric slide (32), and the control chip (23) controls the arc-shaped electric slide (32), the visual camera (33) and the fill light (34) to perform actions respectively.

8. The automobile paint wear resistance detection device according to claim 7, characterized in that: The wear analysis unit comprises a memory (41) and a computer (42), wherein the memory (41) communicates with a visual camera (33) to exchange visual picture data, and the computer (42) analyzes the wear of each visual picture based on a deep learning model and outputs a wear resistance detection score of a target automobile paint surface (100).

9. A method for detecting the wear resistance of automobile paint, characterized in that: The automobile paint wear resistance detection device according to claim 8 is used to perform physical friction simulation on the target automobile paint surface (100), and visual detection and analysis are performed to output a wear resistance detection score, comprising the following steps: Step 1, select the type of friction plate and install the friction plate on the friction plate mounting block; Step 2, setting the pressure parameter, the friction speed parameter and the friction time parameter through the touch screen (21); Step 3, the friction simulation controller controls the longitudinal electric moving mechanism (12) and the transverse electric moving mechanism (13) to be linked, so that the detachable friction plate assembly (17) is located at the center of the target automobile paint surface (100), controls the servo electric cylinder (15) to apply a set pressure, and controls the transverse electric moving mechanism (13) to perform a linear reciprocating motion at a set friction speed and for a set friction time; Step 4: After step 3 is completed, the friction simulation controller controls the longitudinal electric moving mechanism (12) to move the detachable friction plate assembly (17) away from the target automobile paint surface (100); Step 5, the friction simulation controller controls the slider movement of the arc-shaped electric slide (32), from the leftmost side to the rightmost side, and stops for one to two seconds every (20)° arc, and the camera (33) and the fill light (34) are turned on to shoot the visual picture of the current angle. After shooting the visual pictures of all angles, all the visual pictures of the angles are packaged and sent to the wear analysis unit; Step 6: The wear analysis unit analyzes the wear of each visual image based on the deep learning model, and outputs the wear resistance detection score of the target automobile paint surface (100).

10. A method for detecting the wear resistance of automobile paint according to claim 9, characterized in that: In step 6, a convolutional neural network is used to identify and classify different types of wear; and the degree of wear is quantified by analyzing image features. The formula is: Where W is the wear amount, PX a is the number of pixels in the worn area identified by the deep learning model, PX all is the total number of pixels; The formula for the wear test score is: Where WL is the wear resistance test score, W max is the maximum wear of all visual images, W av is the mean wear value of all visual images, S is the wear threshold, α and β are weight coefficients, α is 0.6, and β is 0.4.

Images