Precise copper material surface defect identification method and system based on optical online detection

Abstract

The invention relates to the technical field of defect detection, in particular to a precise copper material surface defect identification method and system based on optical online detection.The method comprises the steps that a plurality of light source devices are arranged on the surface of a precise copper material, and a multi-light-source illumination system is constructed; dividing the surface of the precise copper material into a plurality of detection areas, and installing an industrial camera in each detection area; collecting surface images of the precise copper material through a plurality of industrial cameras, and adding a timestamp to each surface image to obtain a copper material surface image; the collected copper material surface image is transmitted to a data processing center for feature processing, and a copper material surface feature image is obtained; and constructing a surface defect identification model, and inputting the copper material surface feature image into the surface defect identification model to obtain a copper material surface defect identification result. According to the invention, the hysteresis and inconvenience of off-line detection are effectively solved, the detection precision and efficiency are greatly improved, and the risks of missing detection and false detection are effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of defect detection, and in particular to a method and system for identifying surface defects of precision copper materials based on optical online detection. Background Art

[0002] Precision copper is an important material widely used in electronics, communications, automobiles, construction and other fields. It has good electrical conductivity, thermal conductivity, corrosion resistance and mechanical strength, and can meet the needs of modern industry for high-performance materials. In modern industrial production, with the advancement of science and technology, the production technology of copper materials is constantly improving, and the market's quality requirements for copper materials are getting higher and higher, especially for precision copper materials. The surface quality directly affects its performance in various applications.

[0003] Traditional copper surface defect detection mainly relies on manual inspection or offline inspection. Manual inspection is inefficient and easily affected by subjective factors, resulting in frequent missed detection or false detection. Especially when facing small or complex surface defects, it is difficult to meet the requirements of accuracy and speed. In addition, due to its inherent limitations, traditional detection methods are difficult to achieve effective detection when the production line is running at high speed. Therefore, a precision copper surface defect identification method based on optical online detection has become a research direction that needs to be studied at present. Summary of the invention

[0004] The present invention provides a method and system for identifying surface defects of precision copper materials based on optical online detection, which can effectively solve the problems in the background technology.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is: A method for identifying surface defects of precision copper materials based on optical online detection, the method comprising: Arrange several light source devices on the surface of the precision copper material to construct a multi-light source lighting system; Dividing the surface of the precision copper material into a number of detection areas, and installing an industrial camera in each of the detection areas; Based on the multi-light source lighting system, a surface image of the precision copper material is collected by a plurality of the industrial cameras, and a timestamp is added to each surface image to obtain a surface image of the copper material; The collected copper material surface image is transmitted to a data processing center for feature processing to obtain a copper material surface feature image; A surface defect recognition model is constructed, and the copper material surface feature image is input into the surface defect recognition model to obtain a copper material surface defect recognition result.

[0006] Furthermore, a plurality of light source devices are arranged on the surface of the precision copper material to construct a multi-light source lighting system, including: Determine the layout position of the light source device, and perform initial calibration on the brightness and angle of the light source device; Install an ambient light sensor to collect light intensity data in the detection environment; Performing quality analysis on the copper material surface image captured by the industrial camera to obtain image quality data; The brightness and angle of the light source device are adjusted in combination with the light intensity data and the image quality data.

[0007] Furthermore, it includes: when collecting the surface image of the precision copper material, controlling the shooting frequency of the industrial camera through an encoder or a sensor.

[0008] Furthermore, it includes: transmitting the collected surface image of the copper material to a data processing center via a high-speed data transmission interface.

[0009] Furthermore, the collected copper material surface image is transmitted to a data processing center for feature processing to obtain a copper material surface feature image, including: Filtering the copper material surface images of the same precision copper material according to the timestamp; Splicing a plurality of copper material surface images of the same precision copper material by an image splicing algorithm to obtain a complete copper material surface image; Preprocessing the complete image of the copper material surface; Feature extraction is performed on the preprocessed complete image of the copper material surface to obtain a copper material surface feature image.

[0010] Furthermore, a plurality of copper material surface images of the same precision copper material are spliced ​​by an image splicing algorithm to obtain a complete copper material surface image, including: Extracting feature points from overlapping areas of a plurality of copper material surface images of the same precision copper material; Matching the feature points of a plurality of the copper material surface images to obtain a feature point matching result; Calculating a transformation matrix between the copper material surface images according to the feature point matching results, and performing a geometric transformation on the copper material surface images according to the transformation matrix; The overlapping areas of the copper material surface images after the geometric transformation are fused to obtain a complete copper material surface image.

[0011] Furthermore, the method further comprises: after matching the feature points of a plurality of copper material surface images, eliminating wrong matching pairs by using a random sampling consistency algorithm.

[0012] A precision copper surface defect recognition system based on optical online detection, the system comprising: Lighting system construction module, arrange several light source devices on the surface of precision copper materials to build a multi-light source lighting system; A detection area division module divides the surface of the precision copper material into a number of detection areas, and an industrial camera is installed in each of the detection areas; A surface image acquisition module, based on the multi-light source lighting system, acquires the surface image of the precision copper material through a plurality of the industrial cameras, and adds a timestamp to each of the surface images to obtain the copper material surface image; A feature processing module transmits the collected copper material surface image to a data processing center for feature processing to obtain a copper material surface feature image; The defect recognition module constructs a surface defect recognition model, inputs the copper material surface feature image into the surface defect recognition model, and obtains the copper material surface defect recognition result.

[0013] Furthermore, the lighting system building module includes: A light source device layout unit, which determines the layout position of the light source device and performs initial calibration on the brightness and angle of the light source device; A light intensity data acquisition unit is installed with an ambient light sensor, and collects light intensity data in the detection environment through the ambient light sensor; An image quality analysis unit performs quality analysis on the copper material surface image captured by the industrial camera to obtain image quality data; The light source device adjustment unit adjusts the brightness and angle of the light source device in combination with the light intensity data and the image quality data.

[0014] Furthermore, the feature processing module includes: An image screening unit, screening the copper material surface image of the same precision copper material according to the timestamp; An image stitching unit stitches a plurality of copper material surface images of the same precision copper material through an image stitching algorithm to obtain a complete copper material surface image; An image preprocessing unit, for preprocessing the complete image of the copper material surface; The feature extraction unit extracts features from the preprocessed complete image of the copper material surface to obtain a copper material surface feature image.

[0015] The technical solution of the present invention can achieve the following technical effects: It effectively solves the lag and inconvenience of offline detection, realizes efficient and real-time online detection, greatly improves detection accuracy and efficiency, effectively reduces the risk of missed detection and false detection, and ensures the stability and consistency of copper quality.

[0016] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flow chart of a method for identifying surface defects of precision copper materials based on optical online detection; Figure 2 A schematic diagram of the process of constructing a multi-light source lighting system; Figure 3 This is a schematic diagram of the structure of the precision copper surface defect recognition system based on optical online detection. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0021] Embodiment 1 like Figure 1 As shown, a method for identifying surface defects of precision copper materials based on optical online detection includes: S1: Arrange several light source devices on the surface of precision copper materials to build a multi-light source lighting system; Specifically, in optical inspection systems, ambient light can interfere with image acquisition, resulting in common problems such as shadows, uneven lighting, and excessive reflections. In order to ensure that the image quality is high enough, a multi-angle lighting method can be used. Multiple light sources are arranged around the copper surface to illuminate the copper surface from different angles. The brightness and angle of the light sources can be adjusted independently, so different light source combinations can be activated as needed. In this way, problems such as shadows caused by a single light source can be eliminated, and the entire surface area can be fully illuminated. The system can dynamically adjust the brightness and angle of the light source according to changes in light in the inspection environment to ensure stable quality of the acquired image.

[0022] S2: Divide the surface of the precision copper material into several inspection areas, and install an industrial camera in each inspection area; In this embodiment, for the shooting efficiency of the industrial camera, the copper surface can be divided into several detection areas according to the width, length and speed of the production line of the copper material, and an industrial camera is installed in each area. The size of each area is determined by the field of view and resolution of the industrial camera, ensuring that the image taken by each camera can completely cover the surface within the area. In order to ensure that the boundaries of the areas photographed by each camera are seamlessly connected, the area photographed by each camera should contain a small amount of overlapping parts with adjacent areas, and the overlapping parts can be used for later splicing to achieve image continuity.

[0023] Partitioning and photographing precision copper materials can reduce the load of a single camera, so that each industrial camera is only responsible for capturing images within a certain range, thereby improving the image resolution and detail capture capabilities. At the same time, partitioning allows multiple industrial cameras to work simultaneously, thus speeding up the inspection process and ensuring full coverage even at high production line speeds. It should be noted that in order to ensure accurate subsequent image stitching, each industrial camera needs to be accurately calibrated during installation, including adjustment of camera position, angle, focal length, and calibration of overlapping areas to ensure spatial alignment and consistency of the image. In addition, during the production process, the camera's field of view may shift slightly due to vibration, temperature changes, and other factors. Dynamic calibration can be performed regularly or in real time to maintain the accuracy of image stitching.

[0024] S3: Based on the multi-light source lighting system, the surface image of the precision copper material is collected by several industrial cameras, and a timestamp is added to each surface image to obtain the surface image of the copper material; S4: transmitting the collected copper material surface image to a data processing center for feature processing to obtain a copper material surface feature image; The copper surface images captured by industrial cameras may contain noise or light and shadow effects. The image quality can be improved through image preprocessing steps, such as filtering, denoising, contrast enhancement, etc. Commonly used techniques include Gaussian filtering, edge detection, histogram equalization, etc.; when performing feature extraction, multi-dimensional feature extraction methods can be introduced. In addition to conventional geometric features, texture features, grayscale features, etc. can also be combined to improve the generalization ability of the model.

[0025] S5: construct a surface defect recognition model, input the copper material surface feature image into the surface defect recognition model, and obtain the copper material surface defect recognition result.

[0026] Specifically, to build a surface defect recognition model, you can choose a convolutional neural network model, a semantic segmentation model, a random forest, etc. By building a surface defect recognition model, inputting the feature image into the model and obtaining the defect recognition result, it is the core of realizing automated detection. Machine learning or deep learning models can learn various defect characteristics through training and have strong generalization capabilities. Using machine learning or deep learning models, you can efficiently identify possible defects on the surface of copper materials.

[0027] Through the present invention, the lag and inconvenience of offline detection are effectively solved, efficient and real-time online detection is achieved, the detection accuracy and efficiency are greatly improved, the risk of missed detection and false detection is effectively reduced, and the stability and consistency of copper quality are ensured.

[0028] In order to obtain high-quality precision copper surface images, such as Figure 2 As shown, several light source devices are arranged on the surface of the precision copper material to construct a multi-light source lighting system, including: A1: Determine the layout of the light source equipment and perform initial calibration of the brightness and angle of the light source equipment; In this step, the specific installation position of the light source can be planned according to the size and shape of the detection area. The light source should be evenly distributed around the copper surface to reduce the impact of shadows and uneven lighting. A surround layout, multi-angle layout or top layout can be used to cover all detection areas on the copper surface. When the system is started, the initial calibration of the light source brightness and angle can be performed. The best combination of light source brightness and angle can be determined based on the image quality of the reference sample. This initial calibration data will serve as the basis for dynamic adjustment in the subsequent detection process.

[0029] A2: Install an ambient light sensor to collect light intensity data in the detection environment; A3: Perform quality analysis on the copper surface images collected by the industrial camera to obtain image quality data; A4: Combine the light intensity data and image quality data to adjust the brightness and angle of the light source equipment.

[0030] Specifically, when collecting light intensity data in the detection environment, suitable ambient light sensors can be selected, such as photodiodes, photoresistors, etc., to monitor the ambient light intensity in the detection area. The sensors should be placed near the light source and the copper surface to monitor changes in ambient light in real time. For copper surface images, image processing technology can be used to analyze image exposure, reflective areas, shadow areas, contrast and other information. The system needs to combine ambient light and image feedback, and decide how to adjust the brightness and angle through preset rules. The system performs closed-loop control based on ambient light sensors and image quality feedback. Adjustments to the brightness and angle of the light source will cause changes in image quality. The system continuously analyzes new image feedback and dynamically corrects the light source settings to ensure that the image quality gradually approaches the optimal state after each adjustment.

[0031] As a preferred embodiment of this embodiment, when collecting the surface image of the precision copper material, the shooting frequency of the industrial camera is controlled by an encoder or a sensor.

[0032] On the production line, copper materials usually move continuously at a certain speed. To ensure that image acquisition is synchronized with the movement of the copper materials, the camera's shooting frequency and timing can be controlled by an encoder or sensor. This can avoid image blur or loss caused by the mismatch between the copper material's movement speed and the camera's shooting frequency. The sensor or encoder can directly trigger the camera to shoot. Whenever the encoder detects that the copper material has moved a specified distance, or the sensor detects that the copper material is at a certain critical position, the camera is triggered to shoot. This can ensure the accuracy and consistency of image acquisition and avoid missing or repeated shooting. When the production line speed changes, the encoder or sensor can also automatically adjust the camera's shooting frequency so that the shooting timing is synchronized with the movement of the copper material, ensuring that the captured image can still fully cover the copper surface.

[0033] In order to ensure that the copper material surface image can be quickly and stably transmitted to the data processing center, the collected copper material surface image is transmitted to the data processing center through a high-speed data transmission interface.

[0034] As a preferred embodiment of this invention, on the one hand, in industrial production lines, especially in high-speed production environments, images of the copper surface need to be transmitted to the data processing center as quickly as possible for processing and analysis to ensure that defects can be detected and responded to in real time. Therefore, data transmission can select high-speed data transmission interfaces such as GigE, USB3.0, CoaXPress, etc. to ensure that data from the camera to the processing unit can be transmitted at the millisecond level. Low-latency and high-bandwidth interfaces can effectively avoid data congestion, thereby supporting real-time processing. On the other hand, industrial cameras usually generate high-resolution images. In particular, in order to detect tiny defects on the surface of precision copper materials, high-resolution or multi-frame capture may be required. The amount of data in each frame is large. The use of high-speed transmission interfaces can ensure that the image is transmitted to the processing center in a short time, thereby avoiding data accumulation or delay.

[0035] On the basis of the above embodiment, the collected copper material surface image is transmitted to the data processing center for feature processing to obtain the copper material surface feature image, including: B1: Filter the copper surface images of the same precision copper material according to the timestamp; B2: stitching several copper material surface images of the same precision copper material through an image stitching algorithm to obtain a complete copper material surface image; Specifically, the production line is equipped with multiple cameras to simultaneously shoot different parts of the same piece of copper, and the shooting of these cameras is synchronized. Therefore, the images taken at the same time point belong to the same piece of copper. The image stitching algorithm can stitch the surface images belonging to the same piece of copper into a complete surface image.

[0036] B3: Preprocess the complete image of the copper surface; B4: Extract features from the pre-processed complete image of the copper material surface to obtain a copper material surface feature image.

[0037] In order to remove noise in the image and enhance key information, the complete image of the copper surface can be preprocessed, including denoising, contrast enhancement, geometric correction, etc. After image preprocessing, important information related to copper surface defects can be extracted from the preprocessed image through edge detection, texture analysis and other methods, including edge, texture, shape and other information; by splicing first and then preprocessing, it can be ensured that the image of the entire copper surface has global consistency during preprocessing. For example, if the brightness or contrast is found to be uneven after splicing, the entire spliced ​​image can be uniformly adjusted to avoid local inconsistencies.

[0038] On the basis of the above-mentioned embodiment, several copper material surface images of the same precision copper material are spliced ​​by an image splicing algorithm to obtain a complete copper material surface image, including: B21: Extract feature points from overlapping areas of several copper material surface images of the same precision copper material; B22: Matching feature points of several copper material surface images to obtain feature point matching results; In this step, feature points can be extracted from the images in the overlapping areas through algorithms such as SIFT, SURF and ORB, including corner points, edge points, local textures, etc. Once the feature points are extracted, the feature points in different images can be matched. By calculating the descriptors of the feature points, the Euclidean distance or other distance metrics can be used to match the feature point pairs in adjacent images.

[0039] B23: Calculate the transformation matrix between the copper material surface images according to the feature point matching results, and perform geometric transformation on the copper material surface images according to the transformation matrix; Specifically, based on the feature point matching results, the transformation matrix between images is calculated, such as the homography matrix, affine transformation matrix, etc. This transformation matrix can transform the coordinate space of one image to the coordinate space of another image to align their coordinate spaces. Among them, geometric transformations include translation, rotation, scaling, perspective transformation, etc. The transformed image will partially overlap with the adjacent image, preparing for subsequent stitching.

[0040] B24: Fusing the overlapping areas of the copper material surface images after geometric transformation to obtain a complete copper material surface image.

[0041] During the stitching process, there will be overlapping areas between adjacent images. In order to make the transition area seamless, it is usually necessary to fuse the overlapping areas. The fusion methods are: (1) Weighted fusion: In overlapping areas, the weighted averaging method can be used to fuse images. For example, linear weighted averaging is used to weight the pixel values ​​in the overlapping areas by distance, so that the area closer to the center of a certain image has a larger weight. This can smooth the transition area and reduce visual abruptness.

[0042] (2) Gradient blending: Use gradient blending to gradually blend pixels in overlapping areas, reduce color and brightness differences, and avoid obvious boundaries at the joints of images.

[0043] (3) Laplacian pyramid fusion: The image is decomposed through a multi-scale Laplacian pyramid and fused at each scale. Finally, the results are synthesized to better handle stitching areas with large brightness differences.

[0044] During the stitching process, images from different cameras or shooting conditions may have differences in brightness and color. Therefore, after the final stitching is completed, the color and brightness of the entire image can be balanced. Technologies such as global histogram matching or local brightness equalization can be used to make the entire image look unified and harmonious.

[0045] In order to improve the stitching accuracy of copper surface images, after matching the feature points of several copper surface images, the wrong matching pairs are eliminated through a random sampling consistency algorithm.

[0046] Specifically, in the feature point matching process, some incorrect matching pairs may occur due to image noise, lighting changes, surface reflections, texture repetition, etc. The incorrect matching pairs will affect the stitching accuracy of the image and may cause problems such as misalignment and distortion in the stitching results. Therefore, a random sampling consistency algorithm can be used to eliminate incorrect matching pairs and find a set of matching point pairs that best fit the transformation model. The incorrect matching pairs that do not fit the model will be eliminated, leaving high-confidence matching results.

[0047] Embodiment 2: like Figure 3 As shown, the precision copper surface defect recognition system based on optical online detection includes: Lighting system construction module, arrange several light source devices on the surface of precision copper materials to build a multi-light source lighting system; The detection area division module divides the surface of the precision copper material into several detection areas and installs an industrial camera in each detection area; The surface image acquisition module, based on a multi-light source lighting system, uses several industrial cameras to acquire the surface image of the precision copper material, and adds a timestamp to each surface image to obtain the copper material surface image; The feature processing module transmits the collected copper material surface image to the data processing center for feature processing to obtain the copper material surface feature image; The defect recognition module builds a surface defect recognition model, inputs the copper material surface feature image into the surface defect recognition model, and obtains the copper material surface defect recognition result.

[0048] The above-mentioned adjustment system in the present invention can effectively realize the precision copper material surface defect identification method based on optical online detection, and the technical effect that can be achieved is as described in the above-mentioned embodiment, which will not be repeated here.

[0049] Further, the lighting system building blocks include: A light source device layout unit determines the layout position of the light source device and performs initial calibration on the brightness and angle of the light source device; A light intensity data acquisition unit is installed with an ambient light sensor to collect light intensity data in the detection environment; An image quality analysis unit performs quality analysis on the copper surface image collected by the industrial camera to obtain image quality data; The light source device adjustment unit adjusts the brightness and angle of the light source device in combination with the light intensity data and the image quality data.

[0050] Furthermore, the feature processing module includes: An image screening unit screens the copper surface images of the same precision copper material according to the timestamp; An image stitching unit stitches several copper material surface images of the same precision copper material through an image stitching algorithm to obtain a complete copper material surface image; An image preprocessing unit is used to preprocess the complete image of the copper material surface; The feature extraction unit extracts features from the preprocessed complete image of the copper material surface to obtain a feature image of the copper material surface.

[0051] Similarly, the above-mentioned optimization schemes for the system can also respectively achieve the optimization effects corresponding to the method in Example 1, which will not be repeated here.

[0052] Although the present application has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present application. Accordingly, this specification and the accompanying drawings are merely exemplary illustrations of the present application as defined herein, and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. A method for identifying surface defects of precision copper materials based on optical online detection, characterized in that: The method comprises: Arrange several light source devices on the surface of the precision copper material to construct a multi-light source lighting system; Dividing the surface of the precision copper material into a number of detection areas, and installing an industrial camera in each of the detection areas; Based on the multi-light source lighting system, a surface image of the precision copper material is collected by a plurality of the industrial cameras, and a timestamp is added to each surface image to obtain a surface image of the copper material; The collected copper material surface image is transmitted to a data processing center for feature processing to obtain a copper material surface feature image; A surface defect recognition model is constructed, and the copper material surface feature image is input into the surface defect recognition model to obtain a copper material surface defect recognition result.

2. The method for identifying surface defects of precision copper materials based on optical online detection according to claim 1 is characterized in that: The method comprises arranging a plurality of light source devices on the surface of the precision copper material to construct a multi-light source lighting system, including: Determine the layout position of the light source device, and perform initial calibration on the brightness and angle of the light source device; Install an ambient light sensor to collect light intensity data in the detection environment; Performing quality analysis on the copper material surface image captured by the industrial camera to obtain image quality data; The brightness and angle of the light source device are adjusted in combination with the light intensity data and the image quality data.

3. The method for identifying surface defects of precision copper materials based on optical online detection according to claim 1, characterized in that: include: When collecting the surface image of the precision copper material, the shooting frequency of the industrial camera is controlled by an encoder or a sensor.

4. The method for identifying surface defects of precision copper materials based on optical online detection according to claim 1, characterized in that: include: The collected copper material surface image is transmitted to a data processing center via a high-speed data transmission interface.

5. The method for identifying surface defects of precision copper materials based on optical online detection according to claim 1, characterized in that: The collected copper material surface image is transmitted to a data processing center for feature processing to obtain a copper material surface feature image, including: Filtering the copper material surface images of the same precision copper material according to the timestamp; Splicing a plurality of copper material surface images of the same precision copper material by an image splicing algorithm to obtain a complete copper material surface image; Preprocessing the complete image of the copper material surface; Feature extraction is performed on the preprocessed complete image of the copper material surface to obtain a copper material surface feature image.

6. The method for identifying surface defects of precision copper materials based on optical online detection according to claim 5 is characterized in that: The plurality of copper material surface images of the same precision copper material are spliced ​​by an image splicing algorithm to obtain a complete copper material surface image, including: Extracting feature points from overlapping areas of a plurality of copper material surface images of the same precision copper material; Matching the feature points of a plurality of the copper material surface images to obtain a feature point matching result; Calculating a transformation matrix between the copper material surface images according to the feature point matching results, and performing a geometric transformation on the copper material surface images according to the transformation matrix; The overlapping areas of the copper material surface images after the geometric transformation are fused to obtain a complete copper material surface image.

7. The method for identifying surface defects of precision copper materials based on optical online detection according to claim 6 is characterized in that: Also includes: After matching the feature points of a plurality of copper material surface images, incorrect matching pairs are eliminated through a random sampling consistency algorithm.

8. A precision copper surface defect recognition system based on optical online detection, characterized in that: The system comprises: Lighting system construction module, arrange several light source devices on the surface of precision copper materials to build a multi-light source lighting system; A detection area division module divides the surface of the precision copper material into a number of detection areas, and an industrial camera is installed in each of the detection areas; A surface image acquisition module, based on the multi-light source lighting system, acquires the surface image of the precision copper material through a plurality of the industrial cameras, and adds a timestamp to each of the surface images to obtain the copper material surface image; A feature processing module transmits the collected copper material surface image to a data processing center for feature processing to obtain a copper material surface feature image; The defect recognition module constructs a surface defect recognition model, inputs the copper material surface feature image into the surface defect recognition model, and obtains the copper material surface defect recognition result.

9. The precision copper material surface defect recognition system based on optical online detection according to claim 8 is characterized in that: The lighting system building blocks include: A light source device layout unit, which determines the layout position of the light source device and performs initial calibration on the brightness and angle of the light source device; A light intensity data acquisition unit is installed with an ambient light sensor, and collects light intensity data in the detection environment through the ambient light sensor; An image quality analysis unit performs quality analysis on the copper material surface image captured by the industrial camera to obtain image quality data; The light source device adjustment unit adjusts the brightness and angle of the light source device in combination with the light intensity data and the image quality data.

10. The precision copper material surface defect recognition system based on optical online detection according to claim 8, characterized in that: The feature processing module comprises: An image screening unit, screening the copper material surface image of the same precision copper material according to the timestamp; An image stitching unit stitches a plurality of copper material surface images of the same precision copper material through an image stitching algorithm to obtain a complete copper material surface image; An image preprocessing unit, for preprocessing the complete image of the copper material surface; The feature extraction unit extracts features from the preprocessed complete image of the copper material surface to obtain a copper material surface feature image.

Images