Automobile connector production surface quality detection method

Abstract

The invention relates to an automobile connector production surface quality detection method, and relates to the technical field of automobile connector surface detection, and the method comprises the steps: obtaining a connector surface image through an image collection device; the image preprocessor performs affine transformation on the image to generate a transformed image, and extracts a detail layer image; the filtering unit acts on the detail layer image according to the filtering kernel to generate a filtering image; positioning and determining a suspected abnormal region; combining the two suspected areas to determine a final abnormal range; and the calculation unit constructs a Gaussian distribution model according to the non-abnormal pixel points, and judges whether the connector is qualified or not through comparison with an abnormal region, so that the detection rate is remarkably improved, and meanwhile, relatively high recognition accuracy is kept. And complicated mathematical operation is applied to the image sequence, so that the defect that general static analysis is easily interfered by external noise is overcome, and the overall robustness of the algorithm is enhanced.

Description

Technical Field

[0001] The present application relates to the technical field of automobile connector surface detection, and in particular to a method for detecting the production surface quality of automobile connectors. Background Art

[0002] At present, with the continuous development of the connector industry, the precision requirements for connector mold parts are getting higher and higher, especially for the surface quality of precision connector parts, which has a great impact on the life of the connector. The surface quality of the connector determines the wear resistance and lubrication conditions of the connector. When oxidation, damage, roughness and uneven defects appear on the surface quality during the production process of the connector, the surface layer metal of the part will undergo structural changes, affecting the welding quality, and the wear will be aggravated during subsequent use, and even peeling may occur. Therefore, the surface quality of the connector must be controlled within a certain range during the production process, and the detection of surface quality is crucial.

[0003] Regarding the above-mentioned technologies, several surface defect detection methods commonly used in the industry currently include laser scanning, optical coherence tomography (OCT), X-ray detection, etc. Among them, laser scanning is mainly used to quickly obtain three-dimensional contour information of objects; OCT can penetrate several millimeters deep into the material to form high-quality cross-sectional images; X-rays can penetrate thicker materials to reveal hidden defects. These methods each have their own advantages but also have limitations, such as high cost, slow speed, and difficulty in adapting to a variety of materials.

[0004] Therefore, a surface quality inspection method for automotive connector production is proposed to significantly improve the inspection rate while maintaining high recognition accuracy. Summary of the Invention

[0005] The purpose of this application is to provide a method for detecting the surface quality of automobile connector production to solve the problems raised by the above background technology.

[0006] The present application provides a method for detecting the surface quality of automobile connector production using the following technical solution: comprising the following steps: S1: The image acquisition device obtains the connector surface image; S2: The image preprocessor performs affine transformation on the image, generates a transformed image, and extracts a detail layer image; S3: The filtering unit acts on the detail layer image according to the filter kernel to generate a filtered image; S4: Locating and determining the suspected abnormal area; S5: Combine the two suspected areas to determine the final abnormal range; S6: The computing unit constructs a Gaussian distribution model based on the non-abnormal pixel points and determines whether the connector is qualified by comparing it with the abnormal area.

[0007] By adopting the above technical solution, it is possible to capture the external image of the automobile connector, and quickly obtain the first suspected abnormal area and the second suspected abnormal area. The final abnormal area is determined by the first suspected abnormal area and the second suspected abnormal area, and the surface quality of the connector is accurately judged thereby.

[0008] Preferably, in step S1, the image acquisition module uses a high-speed camera to acquire surface images, and the image acquisition process performs real-time data caching and processing.

[0009] Preferably, in step S2, the image preprocessor further performs denoising on the collected image by combining median filtering and Gaussian filtering.

[0010] The design of the filter kernel in the filtering unit in step S3 takes into account the texture features of the connector surface and uses filter kernels of various angles and sizes.

[0011] Preferably, the positioning and determination of the suspected abnormal area in step S4 further includes the following steps: S41: The suspected abnormal area positioning unit determines a first suspected abnormal area according to the transformed image and the filtered image; S42: Performing wavelet transformation on the detail layer image to obtain a target image, and then reconstructing the image through inverse operation to define a second suspected abnormal area.

[0012] By adopting the above technical solution, the first suspected abnormal area and the second suspected abnormal area are obtained in different ways, and thereby the final abnormal area is quickly determined.

[0013] Preferably, the suspected abnormal area positioning unit in step S41 adopts an image processing unit, and the image processing unit adopts methods such as threshold segmentation and edge detection to locate and determine the first suspected abnormal area.

[0014] Preferably, in step S42, the second suspected abnormal area is transformed into a target image by performing two wavelet transforms on the detail layer image.

[0015] Preferably, the final abnormal range is confirmed in step S5 by finding the intersection of the first suspected abnormal area and the second suspected abnormal area to obtain the final abnormal area.

[0016] Preferably, in step S6, the computing unit uses a support vector machine (SVM) algorithm to classify abnormal and non-abnormal areas.

[0017] Preferably, the detection process is fed back to the production line in real time, supporting automated adjustment of production parameters.

[0018] In summary, this application includes at least one of the following beneficial technical effects: 1. This application uses an image acquisition device to acquire a connector surface image; an image preprocessor performs an affine transformation on the image to generate a transformed image, and extracts a detail layer image; a filtering unit applies a filter kernel to the detail layer image to generate a filtered image; the suspected abnormal area is located and determined; the two suspected areas are combined to determine the final abnormal range; a computing unit constructs a Gaussian distribution model based on non-abnormal pixels and determines whether the connector is qualified by comparing them with the abnormal area. These steps significantly improve the detection rate while maintaining high recognition accuracy. By applying complex mathematical operations to the image sequence, the disadvantage of general static analysis being easily interfered with by external noise is overcome, thereby enhancing the overall robustness of the algorithm. 2. The method of this application provides an image acquisition device that uses a high-speed camera body to efficiently capture images of automotive connectors. Furthermore, an adjustment frame is provided within the image acquisition device. Driven by first and second screws, the camera body's height and orientation can be flexibly adjusted, significantly enhancing the clarity and integrity of the overall image captured of the connector. 3. The image acquisition device of the present application is also provided with an angle adjustment mechanism, that is, when the servo motor is driven, the high-speed camera body docked to the front side of the mounting plate can be rotated and adjusted, and when the first transmission motor is driven, the worm can be driven by the worm gear, and the fixed arms on both sides indirectly realize the left and right swing adjustment of the high-speed camera body, and when the second transmission motor is driven, the connecting arms on both sides can realize the up and down swing of the connecting frame, thereby indirectly realizing the up and down swing adjustment of the high-speed camera body. In this way, the flexible and large-scale adjustment of the shooting angle of the high-speed camera body can be significantly enhanced, further ensuring the integrity and clarity of the connector image acquisition. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flowchart of this application; Figure 2 It is a flowchart for locating and determining the suspected abnormal area; Figure 3 This is a schematic diagram of the structure of the image acquisition device of this application; Figure 4 This is a schematic diagram of the structure of the adjustment frame of this application; Figure 5 This is a schematic diagram of the combined structure of the angle adjustment mechanism and the high-speed camera body of the present application; Figure 6 This is a schematic diagram of the angle adjustment mechanism structure of the present application; Figure 7 This is a schematic diagram of the three-dimensional structure of the angle adjustment mechanism of the present application; Figure 8 It is a schematic diagram of the three-dimensional structure of another part of the angle adjustment mechanism of this application.

[0020] Explanation of the accompanying drawings: 1. bracket; 2. vertical plate; 3. adjustment frame; 31. embedded plate; 32. first screw; 33. first motor; 34. horizontal plate; 35. second screw; 36. second motor; 4. angle adjustment mechanism; 41. servo motor; 42. docking plate; 43. worm; 44. first transmission motor; 45. worm gear; 46. connecting arm; 47. connecting frame; 48. second transmission motor; 49. fixed arm; 410. mounting plate; 5. high-speed camera body. DETAILED DESCRIPTION

[0021] The following is combined with Figure 1 -Attached Figure 8 , further details of this application are given.

[0022] A method for detecting the surface quality of automobile connector production, referring to Figure 1 and Figure 2 , including the following steps: S1: The image acquisition device obtains the connector surface image; In step S1, the image acquisition module uses a high-speed camera to capture the surface image, and the image acquisition process is performed in real time with data caching and processing to ensure that the connector image is clear and complete. S2: The image preprocessor performs affine transformation on the image, generates a transformed image, and extracts a detail layer image; In step S2, the image preprocessor further performs denoising on the captured image by combining median filtering and Gaussian filtering to adapt to different types of surface defect detection. S3: The filtering unit acts on the detail layer image according to the filter kernel to generate a filtered image; Specifically, in steps S2 and S3, an affine transformation is performed on the surface image to obtain a transformed image, a detail layer image of the transformed image is obtained, and the detail layer image is filtered according to a plurality of preset filter kernels of different directional scales to obtain a filtered image.

[0023] The design of the filter kernel in the filtering unit in step S3 takes into account the texture characteristics of the connector surface and uses filter kernels of various angles and sizes. That is, the use of filter kernels of various angles and sizes can better capture defects in different directions, such as scratches and cracks. The multi-angle filter kernel can adapt to texture changes in different directions, and filter kernels of different sizes can detect defects of different sizes, from tiny scratches to large dents. S4: Locating and determining the suspected abnormal area; Specifically, the location determination of the suspected abnormal area in step S4 further includes the following steps: S41: The suspected abnormal area positioning unit determines a first suspected abnormal area according to the transformed image and the filtered image; S42: Performing wavelet transformation on the detail layer image to obtain a target image, and then reconstructing the image through inverse operation to define a second suspected abnormal area.

[0024] Specifically, the first suspected abnormal region and the second suspected abnormal region are obtained in different ways, and thereby the final abnormal region is determined quickly and accurately.

[0025] Among them, in step S41, the suspected abnormal area positioning unit adopts an image processing unit, and the image processing unit adopts threshold segmentation, edge detection and other methods to locate and determine the first suspected abnormal area to ensure accurate and rapid positioning of the first suspected abnormal area.

[0026] In step S42, the second suspected abnormal area is transformed into a target image by two wavelet transforms on the detail layer. The two wavelet transforms can enhance the ability to capture subtle defects in the detail layer, especially in the surface inspection of automotive connectors, where tiny defects may affect the performance of the connector. Secondly, the two transformations can reduce noise interference and effectively distinguish between real defects and noise through the decomposition and reconstruction process. In addition, multi-scale analysis helps to locate defects of different sizes and improve the comprehensiveness of the inspection. The inverse operation reconstructed image can retain key information while removing redundancy, making the abnormal area more obvious. S5: Combine the two suspected areas to determine the final abnormal range; The final abnormal range in step S5 is determined by finding the intersection of the first suspected abnormal area and the second suspected abnormal area to obtain the final abnormal area; S6: The computing unit constructs a Gaussian distribution model based on the non-abnormal pixel points and determines whether the connector is qualified by comparing it with the abnormal area.

[0027] Specifically, it is possible to capture external images of automobile connectors, and quickly obtain the first suspected abnormal area and the second suspected abnormal area, determine the final abnormal area through the first suspected abnormal area and the second suspected abnormal area, and thereby accurately judge the surface quality of the connector.

[0028] Among them, in step S6, the computing unit uses the support vector machine (SVM) algorithm to classify abnormal and non-abnormal areas, that is, the computing unit can be combined with the Gaussian model to extract features, and then classified through (SVM) to improve accuracy.

[0029] Specifically, refer to Figure 3The image acquisition device includes a bracket 1, a vertical plate 2, an adjustment frame 3, an angle adjustment mechanism 4 and a high-speed camera body 5. The middle part of the rear end of the bracket 1 is vertically connected to the vertical plate 2 for installation. The front side of the vertical plate 2 is connected to the adjustment frame 3. Through the setting of the adjustment frame 3, the image acquisition position can be easily adjusted. The front side of the adjustment frame 3 is connected to the angle adjustment mechanism 4. Through the setting of the angle adjustment mechanism 4, the image acquisition angle can be flexibly adjusted. The front end of the angle adjustment mechanism 4 is connected to the high-speed camera body 5.

[0030] Reference Figure 4 The adjustment frame 3 includes an embedded plate 31, a first screw 32, a first motor 33, a horizontal plate 34, a second screw 35 and a second motor 36. The embedded plate 31 is vertically embedded in the vertical plate 2, the first screw 32 is vertically installed inside the embedded plate 31, the first motor 33 is connected to the upper end of the first screw 32 and is connected to the middle part of the rear side of the upper end of the bracket 1, the horizontal plate 34 is horizontally installed on the front side of the embedded plate 31 and is threadedly connected to the outer side of the first screw 32, the second screw 35 is horizontally installed inside the horizontal plate 34, and the outer side of the second screw 35 is threadedly connected to the rear end of the angle adjustment mechanism 4, and the left side of the second screw 35 is connected to the second motor 36.

[0031] Reference Figure 5 and Figure 8 , the angle adjustment mechanism 4 set in this application includes a servo motor 41, a docking plate 42, a worm 43, a first transmission motor 44, a worm gear 45, a connecting arm 46, a connecting frame 47, a second transmission motor 48, a fixed arm 49 and a mounting plate 410. The rear end of the servo motor 41 is threadedly docked with the second screw 35, and the front end of the servo motor 41 is connected to the docking plate 42. The middle part of the front side of the docking plate 42 is laterally rotatably connected with the worm 43. The left side of the worm 43 is connected to the output end of the first transmission motor 44. The front side of the worm 43 is meshed with the worm gear 45. Connecting arms 46 are sleeved on the left and right sides of the worm 43, and the connecting arms 46 on both sides are fixedly connected to the connecting frame 47 away from the side of the worm 43. Fixed arms 49 are provided at the upper and lower ends of the connecting frame 47, and the fixed arms 49 on both sides are rotatably connected to the upper and lower ends of the worm gear 45 respectively. The front ends of the fixed arms 49 on both sides are fixedly connected to the mounting plate 410, and the front end of the mounting plate 410 is threadedly fastened and docked with the high-speed camera body 5; Specifically, when capturing the image of the connector, after the connector is transferred to the inside of the bracket 1 along with the conveyor belt, the high-speed camera body 5 can be opened to realize the capture of the upper surface of the connector. When the capture activities on the other side of the connector are to be carried out, it is necessary to drive the first motor 33 provided in the adjustment frame 3 in advance to make the first motor 33 realize the rotation of the first screw 32 connected to the bottom. As the first screw 32 rotates, the threaded rod 32 is connected to the outside of the first screw 32 and embedded in the horizontal plate 34 on the front side of the embedded plate 31, and the vertical movement adjustment is realized accordingly to flexibly adjust the capture height and make the image capture clearer. By driving the second motor 36 provided on the left side of the horizontal plate 34, the second screw 35 connected to the output end can be rotated, so that the angle adjustment mechanism 4 connected to the outside of the second screw 35 can realize the left and right movement adjustment of the high-speed camera body 5 as a whole to flexibly adapt to the image capture activities on the left and right sides of the connector. After the high-speed camera body 5 is moved and adjusted to the left or right side of the connector through the transmission of the first screw 32 and the second screw 35, the servo motor 41 is driven to rotate the angle adjustment mechanism 4 as a whole. In this way, the outer lens end of the high-speed camera body 5 is rotated and adjusted to face the connector to be photographed, thereby achieving efficient capture of images on the left and right sides of the connector, ensuring that the subsequently captured images are clear and complete; By driving the first transmission motor 44, the worm 43 connected to the output end of the first transmission motor 44 can be rotated. As the worm 43 rotates, the worm wheel 45 engaged with the front side of the worm 43 will rotate forward and reverse accordingly. As a result, the fixed arm 49 connected to the upper and lower ends of the worm wheel 45 will realize the left and right swing adjustment of the front side relative to the mounting plate 410, thereby further changing the shooting angle of the high-speed camera body 5, so that the high-speed camera body 5 can be tilted for shooting and capturing. At the same time, when the second transmission motor 48 is driven, the connecting arm 46 will be rotated forward and reverse. As the connecting arm 46 moves, the connecting frame 47 connected to the front side of the connecting arm 46 will swing up and down as a whole. The docking effect of the fixed arm 49 and the mounting plate 410 is again achieved to further adjust the shooting angle of the high-speed camera body 5. In this way, the shooting and capturing angle of the high-speed camera body 5 can be greatly enhanced, so that the high-speed camera body 5 can capture the entire image of the connector more clearly and completely, ensuring that the subsequent image analysis is more standardized and avoiding the occurrence of connector misjudgment problems.

[0032] The examples of this specific embodiment are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Identical components are represented by the same reference numerals. Therefore, any equivalent changes made based on the structure, shape, and principle of this application should be included in the scope of protection of this application.

Claims

1. A method for detecting the surface quality of automobile connector production, characterized in that: The following steps are involved: S1: The image acquisition device obtains the connector surface image; S2: The image preprocessor performs affine transformation on the image, generates a transformed image, and extracts a detail layer image; S3: The filtering unit acts on the detail layer image according to the filter kernel to generate a filtered image; S4: Locating and determining the suspected abnormal area; S5: Combine the two suspected areas to determine the final abnormal range; S6: The computing unit constructs a Gaussian distribution model based on the non-abnormal pixel points and determines whether the connector is qualified by comparing it with the abnormal area.

2. The method for detecting surface quality of automobile connector production according to claim 1, characterized in that: In step S1, the image acquisition module uses a high-speed camera to acquire surface images, and the image acquisition process performs real-time data caching and processing.

3. The method for detecting surface quality of automobile connector production according to claim 1, characterized in that: In step S2, the image preprocessor further performs denoising on the collected image by combining median filtering and Gaussian filtering.

4. The method for detecting surface quality of automobile connector production according to claim 1, characterized in that: The design of the filter kernel in the filtering unit in step S3 takes into account the texture features of the connector surface and uses filter kernels of various angles and sizes.

5. The method for detecting surface quality of automobile connector production according to claim 1, characterized in that: The positioning and determination of the suspected abnormal area in step S4 further includes the following steps: S41: The suspected abnormal area positioning unit determines a first suspected abnormal area according to the transformed image and the filtered image; S42: Performing wavelet transformation on the detail layer image to obtain a target image, and then reconstructing the image through inverse operation to define a second suspected abnormal area.

6. The method for detecting surface quality of automobile connector production according to claim 5, characterized in that: In step S41, the suspected abnormal area positioning unit adopts an image processing unit, and the image processing unit adopts methods such as threshold segmentation and edge detection to locate and determine the first suspected abnormal area.

7. The method for detecting surface quality of automobile connector production according to claim 5, characterized in that: In step S42, the second suspected abnormal area is transformed into a target image by performing two wavelet transforms on the detail layer image.

8. The method for detecting surface quality of automobile connector production according to claim 1, characterized in that: The final abnormal range is determined in step S5 by finding the intersection of the first suspected abnormal area and the second suspected abnormal area to obtain the final abnormal area.

9. The method for detecting surface quality of automobile connector production according to claim 1, characterized in that: In step S6, the computing unit uses a support vector machine (SVM) algorithm to classify abnormal and non-abnormal areas.

10. The method for detecting surface quality of automobile connector production according to claim 1, characterized in that: The detection process is fed back to the production line in real time, supporting automated adjustment of production parameters.

Images