Method for detecting quality of wire product based on machine vision
Patent Information
- Application Number
- CN202511988095.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-26
- Publication Date
- 2026-10-27
- Estimated Expiration
- 2045-12-26
AI Technical Summary
传统的质量检测主要依赖人工目视或简单工具进行，不仅效率低下，而且受人员经验与疲劳度影响，易出现漏检、误判等问题
本发明所述的基于机器视觉的电线成品质量检测方法通过机器视觉与非接触式图像分析，实现对电线端子压接质量的自动化检测，显著提升了检测效率与一致性，克服了人工检测主观性强、易疲劳的问题，其次通过批量处理同一批次电线图像，并计算平均端子压接质量指数，可实现对生产质量的整体把控；在质量指数不达标时，自动归集劣质品数据并生成质量预警报告，为生产过程的追溯与工艺优化提供数据支撑，有利于实现质量控制的闭环管理。
Smart Images
Figure CN121903966B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of image recognition and detection technology, and more specifically, to a method for quality inspection of finished wire products based on machine vision. Background Technology
[0002] In the production of finished electrical wires (such as 4.8 terminal wires), the crimping quality of the terminals directly affects their electrical performance and safety. Traditional quality inspection mainly relies on manual visual inspection or simple tools, which is not only inefficient but also susceptible to problems such as missed inspections and misjudgments due to the influence of personnel experience and fatigue. Although some automated inspection solutions based on image recognition exist, they are mostly focused on the positioning or dimensional measurement of single areas, making it difficult to comprehensively assess the overall crimping status of key areas of the terminal (such as the insulation crimping area, the wire core crimping area, and the buffer transition area). In addition, existing methods usually do not consider the relative positional relationship between the overall wire routing and local areas, and lack automatic identification of the insulation reserve status of the buffer transition area, resulting in inaccurate inspection results and failing to provide detailed data support for process improvement. Therefore, there is an urgent need for an inspection method that can efficiently, comprehensively, and automatically identify and quantitatively evaluate the crimping quality of terminals. Summary of the Invention
[0003] The purpose of this invention is to provide a machine vision-based method for quality inspection of finished wire products, so as to solve the above-mentioned technical problems.
[0004] To achieve the above objectives, the embodiments of this application provide the following technical solutions: On one hand, this application provides a machine vision-based method for quality inspection of finished wire products. The method includes: responding to a camera quality inspection trigger command, a camera positioned directly above the inspection platform captures images of multiple finished wire products from the same batch placed manually on the inspection platform, denoted as a first quality inspection image; performing image segmentation on multiple finished wire products in the first quality inspection image based on a color-based image feature extraction algorithm to obtain a first local inspection area corresponding to each finished wire product, and sequentially evaluating the terminal crimping quality index of the finished wire products corresponding to each first local inspection area; and when the average terminal crimping quality index of multiple finished wire products is lower than a preset threshold, collecting terminal crimping data corresponding to multiple inferior finished wire products, and generating a corresponding quality warning report based on the terminal crimping data corresponding to the inferior finished wire products, wherein the inferior finished wire products are those with a terminal crimping quality index lower than the preset threshold.
[0005] Optionally, the finished wire is a 4.8 terminal wire, comprising: a wire and a 4.8 terminal disposed at one end of the wire, wherein the 4.8 terminal includes an insulation crimping area, a buffer transition area and a core crimping area arranged sequentially along the wire line direction; The terminal crimping data corresponding to the substandard wire products includes insulation crimping area offset parameters, core crimping area offset parameters, and buffer transition area insulation reservation status.
[0006] Optionally, a color-based image feature extraction algorithm is used to segment the multiple finished wires in the first quality inspection pattern to obtain a first local detection area corresponding to each finished wire, including: The color-based image feature extraction algorithm constructs identification lines corresponding to each wire in the first quality inspection pattern, and constructs a first local detection area corresponding to the identification lines according to the length of the identification lines and preset extension parameters. The extension parameters include extension length and extension width. The extension length is 120%-130% of the terminal length, and the extension width is 60%-70% of the terminal width. The identification lines are used to characterize the wires.
[0007] Optionally, the terminal crimping quality index of the finished wire corresponding to each first local inspection area is evaluated, including: Based on the image feature extraction algorithm, the insulation crimping area, the wire core crimping area and the buffer transition area in the first local detection area are identified in sequence, and the standard center line of the wire is constructed at the same time. The offset parameters between the insulation crimping area and the standard center line of the wire, and the offset parameters between the core crimping area and the standard center line of the wire are evaluated respectively. Based on the offset parameters, the corresponding crimping quality sub-index is calculated by a weighted algorithm. The crimping quality sub-index is then corrected based on the insulation reservation state of the buffer transition area, thereby obtaining the terminal crimping quality index of the finished wire corresponding to the first local detection area.
[0008] Optionally, based on image feature extraction algorithms, the insulation crimping area, the wire core crimping area, and the buffer transition area in the first local detection area are identified sequentially, while a standard centerline for the wire is constructed, including: A color-based image feature extraction algorithm identifies the insulation crimping area and the self-locking area, wherein the self-locking area includes the wire core crimping area; The wire core crimping area in the self-locking port area is divided and extracted based on the preset functional division block diagram of the self-locking port area. The area between the wire core crimping area and the insulation crimping area is designated as the buffer transition area; Taking the side of the first rectangle closest to the wire as the starting edge, a first preset length is cut along the identification line and recorded as the standard center line of the wire.
[0009] Optionally, the offset parameters between the insulation crimping area and the standard neutral line of the wire, and the offset parameters between the core crimping area and the standard neutral line of the wire are evaluated. Based on the offset parameters, the corresponding crimping quality sub-index is calculated using a weighted algorithm, including: A first rectangular frame is constructed based on the insulation crimping area, a second rectangular frame is constructed based on the wire core crimping area, and a first center line and a first center point of the first rectangular frame are constructed in the direction of wire extension. Construct the second center line of the second rectangle along the extension direction of the wire, and the second center point of the second rectangle. Calculate the angle between the first center line and the standard center line of the wire, and record it as the first angle. Calculate the angle between the second center line and the standard center line of the wire, and record it as the second angle. Calculate the perpendicular distance between the first center point and the standard center line of the wire, and record it as the first perpendicular distance; Calculate the perpendicular distance between the second center point and the standard center line of the wire, and record it as the second perpendicular distance; The weight values of the vertical distances are configured based on the distribution of the first and second center points relative to the standard center line of the wire. Then, the first crimping quality sub-index corresponding to the insulation crimping area is calculated by weighting the first vertical distance and the first included angle. The second crimping quality sub-index corresponding to the wire core crimping area is obtained by weighted calculation based on the second vertical distance and the second included angle.
[0010] Optionally, the crimping quality sub-index is corrected based on the insulation reservation state of the buffer transition zone, thereby obtaining the terminal crimping quality index of the finished wire corresponding to the first local inspection zone, including: The first crimping quality sub-index and the second crimping quality sub-index are summed, and the presence of a wire encapsulation layer in the buffer transition area is detected based on a color-based image feature recognition algorithm. If the wire encapsulation layer is not present, the sum of the first crimping quality sub-index and the second crimping quality sub-index is corrected based on a preset bias coefficient, and the corrected sum is recorded as the terminal crimping quality index.
[0011] The beneficial effects of this invention are as follows: The machine vision-based quality inspection method for finished wire products described in this invention achieves automated inspection of wire terminal crimping quality through machine vision and non-contact image analysis, significantly improving inspection efficiency and consistency. It overcomes the problems of strong subjectivity and fatigue in manual inspection. Secondly, by batch processing images of the same batch of wires and calculating the average terminal crimping quality index, it can achieve overall control of production quality. When the quality index fails to meet the standard, it automatically collects data on substandard products and generates a quality early warning report, providing data support for production process traceability and process optimization, which is conducive to achieving closed-loop management of quality control.
[0012] Specifically, by constructing a local detection area that matches the wire identification lines, and comprehensively extracting multi-dimensional features such as the center line, included angle, and offset distance of the insulation crimping area and the wire core crimping area, the positional accuracy and morphological compliance of the terminal crimping can be more comprehensively reflected, improving the accuracy of the detection results. Secondly, this invention also introduces the identification of the insulation reservation state of the buffer transition area, and combines it with a preset bias coefficient to correct the quality index, making the evaluation system more complete, helping to discover potential insulation layer fixing defects, and improving the safety and reliability of the product.
[0013] Other features and advantages of the invention will be set forth in the following description, and some of the objects and other advantages may be realized and obtained by means of the structures particularly pointed out in the written description and the drawings. Attached Figure Description
[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0015] Figure 1 This is a schematic diagram a of the finished wire structure described in an embodiment of the present invention; Figure 2 This is schematic diagram b of the finished wire structure described in the embodiment of the present invention; Figure 3 This is a physical image of the finished wire as described in the embodiments of the present invention. Figure 4 This is a schematic diagram of the process for quality inspection of finished wires based on machine vision, as described in an embodiment of the present invention. Detailed Implementation
[0016] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0017] Example 1:
[0018] like Figure 4 As shown in the figure, this embodiment provides a machine vision-based method for inspecting the quality of finished wire products, the method including steps S100, S200 and S300.
[0019] Step S100: In response to the camera quality inspection trigger command, the camera positioned directly above the inspection table captures an image of multiple finished wire products from the same batch that are manually placed on the inspection table. This image is recorded as the first quality inspection image. The finished wire products are manually placed on the inspection table. Figure 3 As shown, the terminal is required to be placed flat on the testing platform to facilitate camera shooting. As can be seen from the figure, the color of the testing platform differs from both the terminal and the wire. The terminal, i.e., the 4.8 terminal, is silver-white, while the wire is black, yellow, red, etc. These color distinctions provide support for subsequent color-based image feature extraction algorithms. The finished wire is a 4.8 terminal wire. Figure 1 - Figure 2 As shown, it includes: a wire and a 4.8 terminal disposed at one section of the wire. The 4.8 terminal includes an insulation crimping area, a buffer transition area, and a core crimping area arranged sequentially along the wire line direction, as shown. Figure 3 As shown, under ideal conditions, the central axis and center point of the insulation crimping area and the wire core crimping area of the terminal are both on the extension line of the wire line. Therefore, the crimping quality of the insulation crimping area can be indirectly characterized based on the offset angle between the central axis of the insulation crimping area and the extension line of the wire line, as well as the angle of deviation of the center point from the vertical length of the extension line of the wire line. The same principle applies to the wire core crimping area, which will not be elaborated here. Step S200: A color-based image feature extraction algorithm constructs identification lines corresponding to each wire in the first quality inspection pattern. These identification lines are used to characterize the placement state of the wire. Based on the length of the identification lines, a first local detection area corresponding to the identification lines is constructed according to preset extension parameters. Taking the identification lines as the center line, the algorithm extends forward and to both sides to obtain the complete placement area of the 4.8 terminal wire. Instead of relying entirely on image feature recognition, only the exposed wire portion is needed to obtain the complete area of the current 4.8 terminal wire through preset image extension parameters. It should be noted that the complete area here does not refer to the entire 4.8 terminal wire, but rather a partial screenshot area including the entire 4.8 terminal wire and a portion of the wire. The extension parameters include extension length and extension width. The extension length is 120%-130% of the terminal length, and the extension width is 60%-70% of the terminal width. Specifically, the algorithm identifies the boundary point between the wire and the insulation crimping area and then extends towards the insulation crimping area, while the width extension extends to both sides by 60%-70% of the terminal width. This results in a square-like screenshot area, which is the first local detection area; Step S300: Analyze the terminal crimping quality index corresponding to each first local detection area in sequence to obtain the average terminal crimping quality index of multiple finished wire products from the same batch. If the average terminal crimping quality index is lower than a preset threshold, collect the terminal crimping data corresponding to multiple substandard finished wire products and generate a corresponding quality warning report based on the terminal crimping data corresponding to the substandard finished wire products. The substandard finished wire products are those whose terminal crimping quality index is lower than the preset threshold. The terminal crimping data corresponding to the substandard finished wire products includes insulation crimping area offset parameters, core crimping area offset parameters, and buffer transition area insulation reservation status. The core of the offset parameters is the center deviation of the insulation crimping area and the core crimping area and the overall centerline offset.
[0020] Secondly, in step S300, the analysis of the terminal crimping quality index corresponding to each first local detection area includes: Step S3101: Identify the insulation crimping area and the self-locking area based on a color-based image feature extraction algorithm. The self-locking area includes the wire core crimping area. Figure 1 and Figure 3 As shown, the self-locking area and the wire core crimping area are connected. Therefore, when using image recognition, these two areas will be identified as a whole. The self-locking area at the front end is not an area to be examined and needs to be removed. At the same time, under the iron core turntable, the image recognition algorithm will also identify the exposed iron core in the buffer transition area as part of the self-locking area, which also needs to be removed. Step S3102: Based on the preset functional division diagram of the self-locking port area, the core crimping area in the self-locking port area is divided and extracted. The specific implementation method is as follows: A self-locking port shape diagram that can be scaled proportionally is set in the program, that is, the functional division diagram. Based on this functional division diagram, the self-locking port area is matched by scaling, and then the self-locking port and part of the exposed iron core in the self-locking port area are removed, and only the core crimping area is retained. The scaling matching benchmark of the functional division diagram is the front end of the self-locking port plus the transition triangle between the self-locking port and the core crimping area. Step S3103: Mark the area between the wire core crimping area and the insulation crimping area as a buffer transition area, such as... Figure 1 - Figure 2 As shown, under standard conditions, the buffer transition zone should include part of the wire insulation layer to ensure the crimping and fixing effect of the insulation crimping zone. Therefore, whether there is an insulation layer in the buffer transition zone is also one of the important evaluation factors of crimping quality. Step S3104: Construct a first rectangular frame based on the insulation crimping area, construct a second rectangular frame based on the wire core crimping area, and construct the first center line and the first center point of the first rectangular frame in the direction of wire extension. Step S3105: Construct the second center line of the second rectangle in the direction of wire extension and the second center point of the second rectangle; Step S3106: Taking the side of the first rectangular frame closest to the wire as the starting side, cut a first preset length along the identification line and record it as the standard center line of the wire. That is, cut a part of the wire screenshot close to the insulation crimping area and then extend it as the standard center line to facilitate the evaluation of the offset of the first rectangular frame and the second rectangular frame. Step S3107: Calculate the angle between the first center line and the standard center line of the wire, and record it as the first angle. Calculate the angle between the second center line and the standard center line of the wire, and record it as the second angle. Step S3108: Calculate the vertical distance between the first center point and the standard center line of the wire, and record it as the first vertical distance; Step S3109: Calculate the perpendicular distance between the second center point and the standard center line of the wire, and record it as the second perpendicular distance; Step S3110: Based on the distribution of the first center point and the second center point relative to the standard center line of the wire, configure the weight value of the vertical distance, and then calculate the first crimping quality sub-index corresponding to the insulation crimping area based on the weighted first vertical distance and the first included angle. There are two distribution situations of the first center point and the second center point relative to the standard center line of the wire. The first case is that the two center points are located on the same side of the standard center line of the wire, which indicates that the overall crimping of the wire is higher or lower than the 4.8 terminal. The second case is that the two center points are located on opposite sides of the standard center line of the wire, which indicates that the wire crimping has a relative misalignment between the insulation crimping area and the wire core crimping area. In this case, the weight of the vertical distance will increase, thereby increasing the attention of the quality evaluation function to this misalignment. Step S311: Calculate the second crimping quality sub-index corresponding to the wire core crimping area based on the weighted calculation of the second vertical distance and the second included angle; Step S312: Summate the first crimping quality sub-index and the second crimping quality sub-index, and detect whether there is a wire encapsulation layer in the buffer transition area based on a color-based image feature recognition algorithm. If not, correct the sum of the first and second crimping quality sub-indexes based on a preset bias coefficient, and record the corrected sum as the terminal crimping quality index. Figure 1 - Figure 2 As shown, under standard conditions, the buffer transition zone should include part of the wire insulation layer to ensure the crimping fixation effect of the insulation crimping zone. Therefore, whether there is an insulation layer in the buffer transition zone is also one of the important evaluation factors of crimping quality. If there is no insulation layer in the buffer transition zone, the terminal crimping quality evaluation index should be further reduced by a preset bias coefficient.
[0021] In this embodiment, automated inspection of wire terminal crimping quality is achieved through machine vision and non-contact image analysis. This significantly improves inspection efficiency and consistency, overcoming the problems of subjective bias and fatigue associated with manual inspection. Specifically, by constructing a local inspection area matching the wire identification lines, multi-dimensional features such as the centerline, angle, and offset distance of the insulation crimping area and the wire core crimping area are comprehensively extracted to fully reflect the positional accuracy and morphological compliance of the terminal crimping, thereby improving the accuracy of the inspection results. Simultaneously, the identification of the insulation reservation state in the buffer transition area is introduced, and the quality index is corrected using a preset bias coefficient, making the evaluation system more complete and helping to discover potential insulation defects, thus improving product safety and reliability. Furthermore, by batch processing images of the same batch of wires and calculating the average terminal crimping quality index, overall control of production quality can be achieved. When quality fails to meet standards, data on substandard products is automatically collected, and quality warning reports are generated, providing data support for production process traceability and optimization, ultimately forming a closed-loop management system for quality control. The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A machine vision-based method for quality inspection of finished wire products, characterized in that, The method includes: In response to the camera quality inspection trigger command, the camera located directly above the inspection table takes pictures of multiple finished wire products from the same batch that are manually placed on the inspection table, and records them as the first quality inspection picture; The color-based image feature extraction algorithm segments the multiple finished wires in the first quality inspection sample to obtain the first local detection area corresponding to each finished wire. The terminal crimping quality index of the finished wire corresponding to each first local detection area is evaluated in turn. If the average terminal crimping quality index of multiple finished wires is lower than a preset threshold, the terminal crimping data corresponding to multiple inferior finished wires are collected. Based on the terminal crimping data corresponding to the inferior finished wires, a corresponding quality warning report is generated. The inferior finished wires are those whose terminal crimping quality index is lower than the preset threshold. Secondly, the terminal crimping quality index of the finished wire corresponding to each first local inspection area is evaluated, including: Based on the image feature extraction algorithm, the insulation crimping area, the wire core crimping area and the buffer transition area in the first local detection area are identified in sequence, and the standard center line of the wire is constructed at the same time. The offset parameters between the insulation crimping area and the standard center line of the wire, and the offset parameters between the core crimping area and the standard center line of the wire are evaluated respectively. Based on the offset parameters, the corresponding crimping quality sub-index is calculated by a weighted algorithm. The crimping quality sub-index is then corrected based on the insulation reservation state of the buffer transition area, thereby obtaining the terminal crimping quality index of the finished wire corresponding to the first local detection area. Secondly, based on image feature extraction algorithms, the insulation crimping area, core crimping area, and buffer transition area in the first local detection area are identified sequentially. Simultaneously, a standard centerline for the wire is constructed, including: A color-based image feature extraction algorithm identifies the insulation crimping area and the self-locking area, wherein the self-locking area includes the wire core crimping area; The wire core crimping area in the self-locking port area is divided and extracted based on the preset functional division block diagram. The area between the wire core crimping area and the insulation crimping area is designated as the buffer transition area; Using the side of the first rectangle closest to the wire as the starting edge, cut a first preset length along the identification line, which is recorded as the standard center line of the wire.
2. The machine vision-based method for inspecting the quality of finished wire products according to claim 1, characterized in that, The finished wire is a 4.8 terminal wire, comprising: a wire and a 4.8 terminal disposed at one end of the wire, wherein the 4.8 terminal includes an insulation crimping area, a buffer transition area and a core crimping area arranged sequentially along the wire line direction; The terminal crimping data corresponding to the substandard wire products includes insulation crimping area offset parameters, core crimping area offset parameters, and buffer transition area insulation reservation status.
3. The machine vision-based method for inspecting the quality of finished wire products according to claim 2, characterized in that, A color-based image feature extraction algorithm segments the multiple finished wires in the first quality inspection sample to obtain the first local detection area corresponding to each finished wire, including: The color-based image feature extraction algorithm constructs identification lines corresponding to each wire in the first quality inspection pattern, and constructs a first local detection area corresponding to the identification lines according to the length of the identification lines and preset extension parameters. The extension parameters include extension length and extension width. The extension length is 120%-130% of the terminal length, and the extension width is 60%-70% of the terminal width. The identification lines are used to characterize the placement state of the wires.
4. The machine vision-based method for inspecting the quality of finished wire products according to claim 1, characterized in that, The offset parameters between the insulation crimping area and the standard neutral line of the wire, and the offset parameters between the core crimping area and the standard neutral line of the wire are evaluated. Based on the offset parameters, the corresponding crimping quality sub-index is calculated using a weighted algorithm, including: A first rectangular frame is constructed based on the insulation crimping area, a second rectangular frame is constructed based on the wire core crimping area, and a first center line and a first center point of the first rectangular frame are constructed in the direction of wire extension. Construct the second center line of the second rectangle along the extension direction of the wire, and the second center point of the second rectangle. Calculate the angle between the first center line and the standard center line of the wire, and record it as the first angle. Calculate the angle between the second center line and the standard center line of the wire, and record it as the second angle. Calculate the perpendicular distance between the first center point and the standard center line of the wire, and record it as the first perpendicular distance; Calculate the perpendicular distance between the second center point and the standard center line of the wire, and record it as the second perpendicular distance; The weight values of the vertical distances are configured based on the distribution of the first and second center points relative to the standard center line of the wire. Then, the first crimping quality sub-index corresponding to the insulation crimping area is calculated by weighting the first vertical distance and the first included angle. The second crimping quality sub-index corresponding to the wire core crimping area is obtained by weighted calculation based on the second vertical distance and the second included angle.
5. The machine vision-based method for inspecting the quality of finished wire products according to claim 4, characterized in that, Based on the insulation reservation state of the buffer transition zone, the crimping quality sub-index is corrected, thereby obtaining the terminal crimping quality index of the finished wire corresponding to the first local inspection zone, including: The first crimping quality sub-index and the second crimping quality sub-index are summed, and the presence of a wire encapsulation layer in the buffer transition area is detected based on a color-based image feature recognition algorithm. If the wire encapsulation layer is not present, the sum of the first crimping quality sub-index and the second crimping quality sub-index is corrected based on a preset bias coefficient, and the corrected sum is recorded as the terminal crimping quality index.