End face material scrap defect detection method and system suitable for pole piece slitting and storage medium

Abstract

The invention provides an end face material scrap defect detection method suitable for pole piece slitting, which comprises the following steps: acquiring an R channel acquisition image of a to-be-detected pole piece slitting end face, and determining a to-be-detected target area; adopting a binarization threshold to extract all foil pixels in the target area, and determining a reference area of the foil; wherein the reference area reflects the main contour of the foil; and retaining the foil pixels outside the reference area, and generating a scrap defect detection result in combination with a preset caliper threshold. The technical problem that in the prior art, the foil scrap defect detection efficiency of the slitting end face of the pole piece is relatively low is effectively solved.

Description

Technical Field

[0001] This application relates to the field of battery electrode inspection technology, specifically to a method, system, and storage medium for detecting end-face material chip defects suitable for electrode cutting. Background Technology

[0002] In the battery manufacturing process, wide battery electrode rolls need to be slit into multiple rolls of the designed width. The inventors discovered that during the slitting process, fine foil debris falls onto the outer coating of the foil at the slit surface, affecting battery production quality. However, existing technologies cannot meet the requirement for rapid detection of foil debris at the slit end face. Summary of the Invention

[0003] In view of this, this application provides a method, system and storage medium for detecting end face chip defects suitable for electrode cutting, so as to solve the technical problem that the prior art cannot quickly and effectively detect foil chip defects on the end face of electrode cutting.

[0004] In a first aspect, this application provides a method for detecting end-face chip defects suitable for electrode cutting, including: Acquire the R-channel image of the end face of the electrode to be inspected and determine the target area to be inspected; All foil pixels within the target area are extracted using a binarization threshold, and a reference area for the foil is determined; wherein, the reference area reflects the main contour of the foil; The foil pixels outside the reference area are retained, and combined with the preset caliper threshold, the debris defect detection result is generated.

[0005] Secondly, this application provides a chip defect detection system suitable for electrode cutting, comprising: The acquisition module is used to acquire the R-channel image of the end face of the electrode to be detected and to determine the target area to be detected. The reference module is used to extract all foil pixels within the target area using a binarization threshold and to determine the reference area of ​​the foil; wherein the reference area reflects the main outline of the foil; The detection module is used to retain foil pixels outside the reference area and generate debris defect detection results by combining them with a preset caliper threshold.

[0006] Thirdly, this application provides a computer device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to realize the end face chip defect detection method of the first aspect suitable for electrode slitting.

[0007] Fourthly, this application provides a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the end-face chip defect detection method suitable for electrode slitting according to the first aspect.

[0008] The method, system, and storage medium for detecting end-face material chip defects suitable for electrode sheet cutting provided in this application have at least the following beneficial effects: By using R-channel images and region focusing, irrelevant background interference is quickly eliminated, directly locking onto chip candidates, resulting in high detection efficiency and eliminating the need for complex preprocessing. This achieves rapid and targeted detection of minute chips, meeting the speed requirements of production lines. The strategy of first extracting all foil pixels and then distinguishing between the main body and chips using a reference region effectively addresses image differences caused by different environments and electrode materials, ensuring the stability and adaptability of the detection method under different production batches and operating conditions. By defining a reference region to represent the main body of the foil, pixels connected to or belonging to the main body are excluded from chip candidates, avoiding misjudging normal, minute foil edge irregularities or image noise as chips, significantly improving detection accuracy and reliability, and reducing the false alarm rate. Attached Figure Description

[0009] Figure 1 A schematic diagram of a method for detecting end-face chip defects suitable for electrode cutting is shown; Figure 2 A schematic diagram of an R-channel image acquisition of the end face of an electrode slicing is shown; Figure 3 A method based on is shown Figure 2 A schematic diagram of separated foil pixels; Figure 4 A method based on is shown Figure 3 A schematic diagram distinguishing between the reference area and the foil pixels outside the area; Figure 5 It shows a kind of corresponding Figure 2 A schematic diagram of foil fragments marked on a color image; Figure 6 A schematic diagram of an end face chip defect detection system suitable for electrode cutting is shown. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, not all of them. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0011] Example 1 See Figure 1 As shown, this application provides a method for detecting end face chip defects suitable for electrode sheet cutting, which includes the following steps.

[0012] S1. Acquire the R-channel image of the end face of the electrode to be detected and determine the target area to be detected; S2. Use binarization thresholding to extract all foil pixels within the target area and determine the reference area of ​​the foil; whereby the reference area reflects the main outline of the foil; S3. Retain foil pixels outside the reference area and combine them with preset caliper thresholds to generate debris defect detection results.

[0013] This application first acquires images of the electrode cutting end face using the R channel, such as... Figure 2 As shown, this highlights the reflectivity of the foil under a specific spectrum, making it easier to distinguish the foil area from the background (such as coating). Based on this, a target area to be detected is determined, focusing on the key location of the slit end face to avoid invalid analysis of the entire image. Subsequently, a binarization thresholding technique is used to segment the pixels within the target area into foil foreground and coating background based on grayscale values, thereby extracting all pixels that may belong to the foil. The separated pixels belonging to the foil are shown below. Figure 3 As shown. See also Figure 4 As shown, these pixels include both the continuous region constituting the main body of the foil, i.e., the reference region, and small debris that has detached from the main body. A reference region is identified and determined that best represents the continuous portion of the foil main body. This reference region reflects the main contour of the foil, such as an ideal rectangular or approximately rectangular area of ​​the foil cross-section after slicing. Finally, all foil pixels extracted from the target area are compared with the reference region, retaining isolated or discrete foil pixels located outside the reference region. These pixel groups typically correspond to detached debris. Combined with preset calibrated thresholds, such as the maximum allowable size and area of ​​the debris, these candidate debris are screened and judged, ultimately generating a clear debris defect detection result, identifying the location and size of the defect. See [link to relevant documentation]. Figure 5 As shown. It should be noted that, Figure 5 The color images are presented solely to more visually demonstrate the foil, coating, and any debris defects marked on the coating.

[0014] This application specifically addresses the detection of foil debris generated during the electrode slitting process. By utilizing R-channel imaging and region focusing, it quickly eliminates irrelevant background interference, directly identifying debris candidates. This high-efficiency detection eliminates the need for complex preprocessing, enabling rapid and targeted detection of minute debris, meeting the speed requirements of production lines. The strategy of first extracting all foil pixels and then distinguishing between the main body and debris using a reference region effectively addresses image differences caused by varying environments and electrode materials (coating color, foil reflectivity), ensuring the stability and adaptability of the detection method across different production batches and operating conditions. By defining a reference region to represent the main body of the foil, pixels connected to or belonging to the main body are excluded from debris candidates. This avoids misjudging normal, minor foil edge irregularities or image noise as debris, significantly improving detection accuracy and reliability while reducing the false alarm rate.

[0015] Furthermore, it also includes: acquiring a G-channel image of the end face of the electrode to be detected after slicing; extracting the coating area based on the G-channel image; and determining the target area to be detected on the R-channel image based on the coating area.

[0016] In addition to acquiring the R-channel image, the G-channel image of the same slit end face is also acquired simultaneously. Because the coating layer in the battery electrode has different grayscale characteristics in the G-channel compared to the foil, it is easier to distinguish. By analyzing and processing the G-channel image, such as thresholding and edge detection, the contour and position information of the coating area on the end face can be extracted more accurately. Since the foil area on the slit end face is usually surrounded or adjacent to the coating area, after determining the coating area, a target area that needs to be carefully inspected for debris can be delineated in its vicinity or at a specific relative position. This target area is then mapped back to the R-channel image with a resolution or features more favorable for foil extraction for subsequent debris detection steps.

[0017] By leveraging complementary information from multiple channels, the coated area is precisely located using the G-channel image, thereby indirectly and accurately pinpointing the exposed foil area most prone to debris adhesion—the target region. This avoids region delineation errors caused by insufficient contrast or interference in a single R-channel image. This method is particularly effective for electrode types where the coating color and foil material have indistinct contrast in the R-channel. The auxiliary positioning using the G-channel ensures reliable identification of the key detection area regardless of the coating color, expanding its applicability. It reduces invalid detection data in non-coated areas, lowers computational load, and avoids false detections caused by interfering pixels in non-coated areas. Precise target region delineation means eliminating more irrelevant background, allowing subsequent binarization and debris analysis to focus on the most relevant image parts, improving the overall processing efficiency and accuracy of the algorithm.

[0018] Furthermore, before determining the reference region, morphological opening operations are used for denoising; the reference region is determined based on the main contour of the foil after denoising.

[0019] After extracting all foil pixels within the target area using the R-channel image, the foil pixel set may contain some noise, such as isolated spots caused by image sensor noise or uneven lighting, or small discontinuous areas formed by slight reflections or contamination on the foil surface. If these noisy pixels are not removed, they may interfere with the determination of the foil's main contour. Therefore, before determining the reference region, a morphological opening operation is performed on the extracted binary image of the foil pixels. The opening operation is a combination of erosion and dilation. Erosion eliminates small protrusions and breaks small connections, thus smoothing object boundaries and breaking narrow connections; the subsequent dilation operation restores the approximate size of the eroded, interest-bearing main body. After the opening operation, most isolated noise points and small burrs are removed, resulting in better connectivity and smoother boundaries in the main foil region. Based on this denoised image, methods such as contour finding and minimum bounding rectangle fitting are used to determine the reference region that best represents the continuous part of the foil's main body.

[0020] Morphological opening operations can filter out noise without significantly altering the shape and area of ​​the main body region, providing a cleaner data foundation for the accurate extraction of the subsequent reference region. After noise removal, the outline of the foil body is clearer and more complete. This makes it less susceptible to interference from irrelevant pixels when determining the reference region, and the obtained reference region more accurately reflects the ideal main outline of the foil cutting end face, reducing reference region offset or deformation caused by noise. This improves the algorithm's tolerance to image quality fluctuations. Even with images of varying quality obtained under different imaging conditions, standard denoising processing can output relatively stable intermediate results, thus ensuring the consistency of the final debris detection results.

[0021] Furthermore, the foil pixels outside the reference area are retained, and combined with the preset gauge threshold, the debris defect detection result is generated, including: for each foil outside the reference area, it is determined whether the length / width / area of ​​the foil is within the preset length / width / area gauge threshold range; if any one of the length / width / area of ​​the foil is not within the length / width / area gauge threshold range, it is determined to be a debris defect.

[0022] After separating the reference region representing the main body of the foil, the remaining independent, connected sets of foil pixels are considered candidate debris objects. For each such candidate object, its geometric characteristics are measured, including: length (e.g., main axis length), width (e.g., minor axis length), and projected area (number of pixels). Simultaneously, one or more sets of caliper thresholds are pre-set according to the quality control standards of the electrode slitting process. For example: maximum allowable length L_max, maximum allowable width W_max, and maximum allowable area A_max. The measured length, width, and area of ​​each candidate debris are checked one by one. If any of the measured values ​​exceeds the corresponding preset threshold (e.g., length > L_max, width > W_max, or area > A_max), the candidate object is determined to be a non-compliant debris defect, ensuring that any abnormal foil particles exceeding the standard in any dimension are captured.

[0023] The definition of debris defects is shifted from subjective visual judgment to objective measurement based on clearly defined dimensional parameters, eliminating inconsistencies caused by subjective human factors and ensuring repeatability and comparability of test results. Preset caliper thresholds can be flexibly adjusted according to different electrode types, quality standards, or slitting processes. For example, stricter thresholds can be set for high-energy-density battery electrodes, enabling refined control of defects. By simultaneously checking length, width, and area, different types of debris can be effectively identified, whether it's long, thin strips exceeding the length limit, large, sheet-like debris exceeding the area limit, or any abnormal protrusions exceeding the width or area limit, ensuring comprehensive testing.

[0024] Furthermore, it also includes: based on a preset merging threshold, determining whether foil materials whose length, width, and area are all within the length / width / area gauge threshold range need to be merged with adjacent foil materials; if merging is required, performing fusion processing using morphological closing operations to obtain the merged foil material, and regenerating the defect detection results of the merged foil material.

[0025] In real-world inspection scenarios, a real debris fragment might be segmented into multiple adjacent, independent small regions in an image (e.g., due to uneven image contrast, gaps in the fragment itself, or attachments). After initial screening, for candidate objects whose size parameters are all within the acceptable threshold range—that is, sets of tiny foil pixels initially determined to be non-defective—further evaluation is conducted to determine if they are spatially close enough to physically belong to the same fragment. To address this, a merging threshold is introduced, such as the maximum allowable distance between the centroids of two regions or the minimum gap between their boundaries. These acceptable small regions are iterated through; if the distance between two regions is found to be less than the merging threshold, they are determined to need to be merged. For multiple regions requiring merging, a morphological closing operation is used. The closing operation is a combination of dilation and erosion. First, dilation is performed on these regions to connect them and merge them into a large connected region; then, erosion is performed to restore an approximate original merged shape and smooth the boundaries, resulting in a merged foil region. For this newly generated fusion region, repeat the above defect detection and judgment process, calculate the new length, width, and area, and compare them with the caliper threshold to generate an updated defect detection result.

[0026] This effectively avoids the situation where a large actual debris is misclassified as multiple acceptable small particles due to image processing, resulting in missed detection. Through the merging operation, the true size of the debris can be restored, ensuring that real defects exceeding the size limit are correctly captured.

[0027] Furthermore, it also includes: calculating the minimum distance between each debris pixel region in the debris defect detection result and the contour boundary of the reference region; if the minimum distance is less than the preset adhesion judgment threshold, it is judged that there is a risk of adhesion between the debris and the foil body, and it is marked as a risk debris to be processed. The merging logic based on spatial proximity conforms to the judgment that should belong to the same object, making the output results of the detection algorithm more consistent with the real physical condition of the electrode end face, improving the reliability and practicality of the detection results. A merging verification mechanism has been added. Even if the object is initially judged to be qualified, it still needs to undergo final size verification after reasonable merging. This forms a rigorous, multi-step screening and verification process, ensuring the accuracy of the final result, avoiding false judgments of qualification due to debris breakage and dispersion, covering the detection scenario of dispersed large debris, and making the detection more comprehensive and without omissions.

[0028] Furthermore, it also includes: real-time recording of debris defect detection results, and generating a time-series distribution map of debris on the slitting end face based on the detection time or the length and position of the slitting roll material; when the distribution frequency or density of debris defects in the distribution map exceeds the preset slitting process allowable threshold within a continuous preset length or time, a slitting equipment parameter adjustment signal or a warning signal is triggered in real time.

[0029] During the inspection process, it's not only crucial to record whether debris is present in each inspection, but more importantly, to link the inspection results (including the number, size, and location of defects) with the spatiotemporal information of the production process. Specifically, there are two main linking methods: one is based on the inspection timestamp; the other is based on the encoder signal indicating the length and position of the electrode sheet slitting roll. Based on this linked data, a real-time debris distribution map of the slitting end face can be generated. This map uses time or roll length as the horizontal axis and the frequency of debris occurrence (e.g., the number of defects per unit length) or distribution density (e.g., the concentration of defects) as the vertical axis, visually displaying the distribution of debris defects across the entire electrode sheet roll. This distribution map is continuously analyzed, and a permissible threshold for the slitting process is set. When it is found that the frequency or density of debris continuously exceeds this threshold within a preset length (e.g., 4 meters) or a preset time (e.g., 1 minute), it indicates an abnormality in the slitting process, and a trend of deteriorating cleanliness has formed. At this point, it's no longer just about reporting defects, but about triggering control signals in real time. For example, sending signals to the slitting machine to adjust parameters (such as reducing slitting speed or adjusting slitting intensity), or sending audible and visual warning signals to the monitoring center to remind operators to intervene and inspect immediately. This elevates single, static defect detection to continuous and trend-based monitoring of the entire slitting production process.

[0030] Upgrading discrete point detection to continuous line monitoring and surface analysis transforms quality control from post-event statistics to real-time intervention, significantly enhancing the intelligence level of the slitting and inspection process. Time-series distribution diagrams are crucial data from the slitting and inspection process, enabling analysis of factors such as the slitting machine's lifespan, the impact of different batches of electrode sheets (different types and thicknesses of foil), and differences in slitting parameters. This provides data-driven decision support for continuous improvement of the slitting process and preventative equipment maintenance. By setting thresholds and automatic triggering mechanisms, quality drift in the electrode sheet slitting process can be detected and stopped promptly, preventing the generation of numerous defective products and ensuring consistently high cleanliness of the output electrode sheet rolls, thus guaranteeing the overall production quality of the battery from the source.

[0031] Example 2 This embodiment provides an end-face chip defect detection system suitable for electrode cutting, which is applied to the end-face chip defect detection method suitable for electrode cutting provided in Embodiment 1 above for illustration. See also Figure 6 As shown, the end face chip defect detection system suitable for electrode slitting provided in this application may include the following multiple modules.

[0032] The acquisition module is used to acquire the R-channel image of the end face of the electrode to be detected and to determine the target area to be detected. The reference module is used to extract all foil pixels within the target area using a binarized threshold and to determine the reference area of ​​the foil; wherein, the reference area reflects the main contour of the foil; The detection module is used to retain foil pixels outside the reference area and, in combination with a preset caliper threshold, generate debris defect detection results.

[0033] The end-face chip defect detection system suitable for electrode cutting provided in this application embodiment can be applied to the end-face chip defect detection method suitable for electrode cutting provided in Embodiment 1 above. For relevant details, please refer to the above method embodiment. Its implementation principle and technical effect are similar, and will not be repeated here.

[0034] It should be noted that the end-face chip defect detection system suitable for electrode slicing provided in this embodiment is only illustrated by the above-described division of functional modules / units. In practical applications, the above functions can be assigned to different functional modules / units as needed, that is, the internal structure of the end-face chip defect detection system suitable for electrode slicing can be divided into different functional modules / units to complete all or part of the functions described above. Furthermore, the implementation method of the end-face chip defect detection method suitable for electrode slicing provided in Embodiment 1 above and the implementation system of the end-face chip defect detection system suitable for electrode slicing provided in Embodiment 2 of this application belong to the same concept. The specific implementation process of the end-face chip defect detection system suitable for electrode slicing provided in Embodiment 2 of this application is detailed in Embodiment 1 above, and will not be repeated here.

[0035] Example 3 One embodiment of this application also provides a computer device, including but not limited to a processor and a memory, wherein the memory stores computer instructions, and the processor executes the computer instructions to implement the method described above.

[0036] One embodiment of this application also provides a computer-readable storage medium for storing a computer program that, when executed by a processor, implements the method described in the above-described method embodiments.

Claims

1. A method for detecting end-face chip defects suitable for electrode sheet cutting, characterized in that, include: Acquire the R-channel image of the end face of the electrode to be inspected and determine the target area to be inspected; All foil pixels within the target area are extracted using a binarization threshold, and a reference region of the foil is determined; wherein, the reference region reflects the main contour of the foil; The foil pixels outside the reference area are retained, and combined with the preset caliper threshold, the debris defect detection result is generated.

2. The method for detecting end-face material chip defects suitable for electrode sheet cutting according to claim 1, characterized in that, Also includes: Acquire the G-channel image of the cut end face of the electrode to be inspected; The coating area is extracted from the image acquired through the G channel, and the target area to be detected is determined on the image acquired through the R channel based on the coating area.

3. The method for detecting end-face material chip defects suitable for electrode sheet cutting according to claim 2, characterized in that: Before determining the reference region, morphological opening operations are used for noise reduction. The reference area is determined based on the main contour of the foil after noise reduction.

4. The method for detecting end-face chip defects suitable for electrode sheet cutting according to claim 3, characterized in that, The process of retaining foil pixels outside the reference area, combined with a preset gauge threshold, generates a debris defect detection result, including: For each foil outside the reference area, determine whether the length / width / area of ​​the foil is within the preset length / width / area caliper threshold range; If any one of the length, width, or area of ​​the foil is outside the specified length, width, or area caliper threshold, it is determined to be a debris defect.

5. The method for detecting end-face material chip defects suitable for electrode sheet cutting according to claim 4, characterized in that, Also includes: Based on a preset merging threshold, determine whether foils whose length, width, and area are all within the specified length / width / area caliper threshold range need to be merged with adjacent foils; If merging is required, a morphological closing operation is used to perform a fusion process to obtain the fused foil, and the defect detection results of the fused foil are regenerated.

6. The method for detecting end-face material chip defects suitable for electrode sheet cutting according to claim 5, characterized in that: For each debris pixel region in the debris defect detection result, calculate the minimum distance between it and the contour boundary of the reference region; If the minimum distance is less than the preset adhesion judgment threshold, it is determined that there is a risk of adhesion between the debris and the foil body, and it is marked as a risk debris to be processed.

7. The method for detecting end-face chip defects suitable for electrode sheet cutting according to claim 6, characterized in that: The results of the debris defect detection are recorded in real time, and a time sequence distribution map of debris on the slit end face is generated based on the detection time or the length and position of the slit roll material. When the distribution frequency or density of debris defects in the distribution map exceeds the preset threshold allowed by the cutting process within a continuous preset length or time, a cutting equipment parameter adjustment signal or a warning signal is triggered in real time.

8. A system for detecting end-face chip defects suitable for electrode sheet cutting, characterized in that, include: The acquisition module is used to acquire the R-channel image of the end face of the electrode to be detected and to determine the target area to be detected. The reference module is used to extract all foil pixels within the target area using a binarization threshold and to determine the reference area of ​​the foil; wherein the reference area reflects the main outline of the foil; The detection module is used to retain foil pixels outside the reference area and generate debris defect detection results by combining them with a preset caliper threshold.

9. A computer device, characterized in that, include: The device includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to implement the end-face chip defect detection method suitable for electrode slitting as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the end-face chip defect detection method suitable for electrode slitting as described in any one of claims 1-7.

Images