A battery piece scratch type detection method and device based on image processing

Abstract

The application discloses a battery piece scratch type detection method and device based on image processing, and relates to the technical field of battery piece production. The method comprises the following steps: in the production process of a battery piece, a to-be-detected image of a to-be-detected battery piece is collected in real time; a defect detection device is used to detect the to-be-detected battery piece, and it is determined whether scratches exist on the surface of the to-be-detected battery piece; if it is determined that scratches exist on the surface of the to-be-detected battery piece, the to-be-detected image is subjected to gray scale processing, and a gray scale degree value of the to-be-detected image is obtained; the mechanical clamp shape profile of the scratches is restored on the to-be-detected image according to the gray scale degree value, and a scratch profile image is obtained; and the scratch type of the to-be-detected battery piece is determined by a correlation coefficient matching method according to the scratch profile image and a preset scratch type standard image. The battery piece image can be processed and analyzed to determine whether the battery piece is scratched and the scratch type, subjective errors in the manual detection process are reduced, the detection efficiency and accuracy are improved, and the labor cost is greatly reduced.

Description

Technical Field

[0001] This invention relates to the field of battery cell inspection technology, specifically to a method and apparatus for detecting scratch types in battery cells based on image processing. Background Technology

[0002] In modern battery manufacturing, the various mechanical processes on the battery cell production line inevitably cause varying degrees of damage to the cells due to the clamping of mechanical fixtures. This directly affects the battery's performance and lifespan. Therefore, accurately identifying the causes of scratches on the battery cells and taking timely measures is of great significance for improving battery quality and production efficiency.

[0003] Existing technical solutions primarily rely on manually observing and collecting images of solar cells, combined with experience, to determine the cause of scratches. While this method can identify scratches to some extent, it heavily depends on the experience and skill of the personnel. Manual observation and judgment are significantly influenced by subjective factors, making misjudgments likely. Furthermore, this method is inefficient, cannot meet the needs of large-scale production, lacks unified standards and criteria, and can lead to significant discrepancies in judgments among different individuals.

[0004] Therefore, those skilled in the art urgently need to find a new technical solution to address the aforementioned problems. Summary of the Invention

[0005] To overcome the problems existing in related technologies, this invention discloses a method and apparatus for detecting scratch types of battery cells based on image processing.

[0006] According to a first aspect of the disclosed embodiments of the present invention, a method for detecting scratch type of battery cells based on image processing is provided, the method comprising:

[0007] During the production process of solar cells, images of the solar cells to be inspected are acquired in real time.

[0008] The defect detection equipment is used to inspect the battery cell to determine whether there are scratches on the surface of the battery cell;

[0009] If it is determined that there are scratches on the surface of the battery cell to be tested, the image to be tested is processed in grayscale to obtain the grayscale value of the image to be tested;

[0010] Based on the grayscale value, the shape and outline of the scratched mechanical fixture are reconstructed on the image to be detected, and a scratch outline image is obtained;

[0011] The scratch type of the battery cell to be tested is determined by the correlation coefficient matching method based on the scratch contour image and the preset scratch type standard image.

[0012] Optionally, determining the scratch type of the battery cell to be tested based on the scratch contour image and a preset scratch type standard image using a correlation coefficient matching method includes:

[0013] The correlation coefficient ρ between the scratch contour image and the standard image for each scratch type is determined using the correlation coefficient calculation formula.

[0014] The formula for calculating the correlation coefficient is: in,

[0015] g(x,y) represents the pixel values ​​of the standard scratch image, and g′(x,y) represents the pixel values ​​of the scratch contour image.

[0016] If the correlation coefficient between the scratch contour image and the scratch type standard image is greater than a preset coefficient threshold, the scratch type corresponding to the scratch type standard image is determined as the scratch type of the battery cell to be tested.

[0017] Optionally, the defect detection equipment is an EL tester, and the real-time acquisition of the image of the battery cell to be inspected includes:

[0018] An industrial camera positioned above the battery cell captures images of the battery cell at a frequency of 10 frames per second.

[0019] The image with the highest resolution is selected as the image to be tested for the battery cell.

[0020] Optionally, if it is determined that there are scratches on the surface of the battery cell to be tested, performing grayscale processing on the image to be tested to obtain the grayscale value of the image to be tested includes:

[0021] If it is determined that there are scratches on the surface of the battery cell to be tested, the image to be tested is processed in grayscale, and the grayscale value is divided into 0-100% to obtain the grayscale value of the image to be tested.

[0022] Optionally, the step of reconstructing the shape outline of the scratched mechanical fixture on the image to be detected based on the grayscale value to obtain a scratch outline image includes:

[0023] The image to be detected is decomposed according to the gray level value, and the area with a gray level value of 0% is determined as the scratch area;

[0024] Based on the outline of the scratched area, the shape and outline of the scratched mechanical fixture are reconstructed on the image to be detected, and a scratch outline image is obtained.

[0025] Optionally, the method further includes:

[0026] Standard images of each scratch type are drawn based on the contact points between the mechanical fixture and the battery cell during the battery cell production process. The scratch types include: dry basket card mark, wet basket card mark, belt mark, suction cup mark, positioning fixture adjustment mark, and boat frame mark.

[0027] Optionally, the method further includes:

[0028] Obtain the total number of battery cells produced within a preset time period and the number of battery cells with each type of scratch.

[0029] The wear level of the corresponding mechanical fixture within a preset time period is determined based on the proportion of each type of scratched battery cell to the total number of battery cells produced.

[0030] According to a second aspect of the disclosed embodiments of the present invention, a battery cell scratch type detection device based on image processing is provided, the device comprising:

[0031] The image acquisition module acquires images of the battery cells to be inspected in real time during the battery cell production process.

[0032] The EL detection module, connected to the image acquisition module, uses a defect detection device to inspect the battery cell to be inspected and determine whether there are scratches on the surface of the battery cell to be inspected.

[0033] A grayscale processing module, connected to the EL detection module, performs grayscale processing on the image to be detected if it is determined that there are scratches on the surface of the battery cell to be detected, and obtains the grayscale value of the image to be detected.

[0034] The contour restoration module is connected to the grayscale processing module. It restores the shape contour of the scratched mechanical fixture on the image to be detected based on the grayscale value, and obtains the scratch contour image.

[0035] The scratch type determination module is connected to the contour restoration module. Based on the scratch contour image and the preset scratch type standard image, it determines the scratch type of the battery cell to be tested by the correlation coefficient matching method.

[0036] Optionally, the scratch type determination module includes:

[0037] The correlation coefficient calculation unit determines the correlation coefficient ρ between the scratch contour image and the standard image for each scratch type using the correlation coefficient calculation formula.

[0038] The formula for calculating the correlation coefficient is: in,

[0039]

[0040] g(x,y) represents the pixel values ​​of the standard scratch image, and g′(x,y) represents the pixel values ​​of the scratch contour image.

[0041] The scratch type determination unit is connected to the correlation coefficient calculation unit. If the correlation coefficient between the scratch contour image and the scratch type standard image is greater than a preset coefficient threshold, the scratch type corresponding to the scratch type standard image is determined as the scratch type of the battery cell to be tested.

[0042] Optionally, the contour restoration module includes:

[0043] The scratch area determination unit decomposes the image to be detected according to the gray level value and determines the area with a gray level value of 0% as the scratch area;

[0044] The scratch contour acquisition unit is connected to the scratch area determination unit. Based on the contour of the scratch area, it reconstructs the shape contour of the scratched mechanical fixture on the image to be detected and acquires the scratch contour image.

[0045] In summary, this invention discloses a method and apparatus for detecting scratch types in battery cells based on image processing. The method includes: acquiring images of the battery cell to be inspected in real time during the battery cell production process; inspecting the battery cell to be inspected using a defect detection device to determine whether scratches exist on the surface of the battery cell; if scratches are found on the surface of the battery cell to be inspected, performing grayscale processing on the image to be inspected to obtain grayscale values; reconstructing the shape and contour of the mechanical fixture with the scratch on the image to be inspected based on the grayscale values ​​to obtain a scratch contour image; and determining the scratch type of the battery cell to be inspected using a correlation coefficient matching method based on the scratch contour image and a preset scratch type standard image. This method can determine whether the battery cell is scratched and the type of scratch by processing and analyzing the battery cell image, reducing subjective errors in the manual inspection process, improving inspection efficiency and accuracy, and significantly reducing labor costs.

[0046] Other features and advantages disclosed in this invention will be described in detail in the following detailed description section. Attached Figure Description

[0047] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation thereof. In the drawings:

[0048] Figure 1 This is a flowchart illustrating an image processing-based method for detecting scratch types in battery cells, according to an exemplary embodiment.

[0049] Figure 2 It is based on Figure 1 This is a schematic diagram illustrating grayscale levels;

[0050] Figure 3 It is based on Figure 1 This diagram illustrates various types of scratches on the surface of a battery cell.

[0051] Figure 4 It is based on Figure 1 A flowchart illustrating a method for analyzing the proportion of scratches on battery cells is shown.

[0052] Figure 5 This is a structural block diagram of an image processing-based battery cell scratch type detection device according to an exemplary embodiment;

[0053] Figure 6 It is based on Figure 5 The diagram shown is a structural block diagram of a scratch type determination module;

[0054] Figure 7 It is based on Figure 5 The diagram shows a structural block diagram of a contour restoration module. Detailed Implementation

[0055] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present disclosure.

[0056] Figure 1 This is a flowchart illustrating an image processing-based method for detecting scratch types in battery cells, according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0057] In step 101, during the production process of the solar cell, the image of the solar cell to be tested is acquired in real time.

[0058] Specifically, an industrial camera positioned above the battery cell captures images of the battery cell at a rate of 10 frames per second; the image with the highest resolution is selected as the image to be tested for the battery cell.

[0059] For example, in an embodiment of the present invention, an industrial camera is installed directly above the battery cell. The industrial camera continuously and at high speed captures images of the battery cells on the production line, and selects the image with the highest resolution from the captured images as the image to be tested, so as to obtain the status information of the battery cell through subsequent image analysis steps. Preferably, the image resolution is 1920*1080.

[0060] In step 102, the battery cell to be tested is inspected using a defect detection device to determine whether there are scratches on the surface of the battery cell.

[0061] Preferably, the defect detection equipment is an EL tester.

[0062] For example, an EL tester, short for Electroluminescent (EL) tester, is a device for detecting internal defects in solar cells or cell modules. This defect detection device is commonly used to detect internal defects, microcracks, fragments, poor soldering, broken grids, and abnormal phenomena in individual cells with different conversion efficiencies in solar cell modules. It is understood that in the embodiments disclosed in this invention, the EL tester performs a "preliminary inspection" of the appearance of the cell to be inspected to determine whether there are defects (i.e., scratches) on the surface of the cell. If the preliminary inspection result shows that there are no defects on the surface of the cell, the processing and analysis process of steps 103-105 below is not performed on the image of the cell to be inspected, and the production process of the cell continues on the production line.

[0063] In step 103, if it is determined that there are scratches on the surface of the battery cell to be tested, grayscale processing is performed on the image to be tested to obtain the grayscale value of the image to be tested.

[0064] For example, as can be seen from step 102 above, after the EL detector performs a "preliminary inspection" on the battery cell to be inspected, if the preliminary inspection result shows that there are defects on the surface of the battery cell, it is necessary to continue processing the image of the battery cell to be inspected in order to obtain the scratch type of the battery cell to be inspected.

[0065] Specifically, if it is determined that there are scratches on the surface of the battery cell to be tested, the image to be tested is processed in grayscale, and the grayscale value is divided into 0-100% to obtain the grayscale value of the image to be tested.

[0066] like Figure 2 As shown, the grayscale values ​​of the image are divided according to the grayscale gradient of 0%, 5%, 10%, 15%, ..., 85%, 90%, 95%, 100%. The image to be detected is processed in grayscale, and the grayscale value is obtained to determine which grayscale gradient range the grayscale value belongs to.

[0067] In step 104, the shape and outline of the scratched mechanical fixture are reconstructed on the image to be detected based on the grayscale value, and the scratch outline image is obtained.

[0068] For example, according to Figure 2It can be seen that the area with a grayscale value of 0% on the image to be inspected is the darkest, and the corresponding area on the battery to be inspected has scratches. The specific location of the area with a grayscale value of 0% on the battery cell to be inspected is determined (it can be represented by the horizontal and vertical coordinates in the battery cell coordinate system), and the division contour is determined based on the contour of the area with a grayscale value of 0%, that is, the shape contour of the mechanical fixture.

[0069] Specifically, the process of reconstructing the shape and outline of the scratched mechanical fixture on the image to be inspected based on the grayscale value and obtaining a scratch outline image includes: disassembling the image to be inspected according to the grayscale value and determining the area with a grayscale value of 0% as the scratch area; and reconstructing the shape and outline of the scratched mechanical fixture on the image to be inspected based on the outline of the scratch area to obtain a scratch outline image.

[0070] For example, after determining the grayscale value of the image to be inspected, the image can be divided into different regions based on the grayscale value. The region with a grayscale value of 0% is the scratch region. The specific location and size of the scratch region are determined based on its coordinate values, and its outline is drawn. It can be understood that the outline of the scratch region is the shape outline of the mechanical fixture that caused the scratch; the image to be inspected that reconstructs the shape outline of the mechanical fixture is the scratch outline image. By comparing and analyzing the scratch outline image with a standard scratch type image, the scratch type of the battery cell to be inspected can be determined.

[0071] In step 105, the scratch type of the battery cell to be tested is determined by correlation coefficient matching based on the scratch contour image and the preset scratch type standard image.

[0072] Specifically, the correlation coefficient ρ between the scratch contour image and the standard image for each scratch type is determined using the correlation coefficient calculation formula; the formula for calculating the correlation coefficient is as follows: in,

[0073] g(x,y) represents the pixel value of the scratch type standard image, and g′(x,y) represents the pixel value of the scratch contour image. If the correlation coefficient between the scratch contour image and the scratch type standard image is greater than a preset coefficient threshold, the scratch type corresponding to the scratch type standard image is determined as the scratch type of the battery cell to be tested.

[0074] For example, after obtaining the scratch contour image through the above steps, the correlation coefficient between the pixel values ​​of the scratch type standard image and the scratch contour image is calculated. If the correlation coefficient is greater than a preset coefficient threshold, it means that the contour on the scratch contour image is the same as the scratch contour of the scratch type standard image, thus determining the scratch type. If the correlation coefficient is less than or equal to the preset coefficient threshold, it means that the contour on the scratch contour image is not the same as the scratch contour of the scratch type standard image. Then, the scratch type standard image corresponding to the next scratch type is compared with the scratch contour image of the battery cell to be tested until the scratch type of the battery cell to be tested is determined.

[0075] Optionally, the method further includes:

[0076] Standard images for each type of scratch are drawn based on the contact points between the mechanical fixtures and the battery cells during the battery cell production process. The scratch types include: dry basket card mark, wet basket card mark, belt mark, suction cup mark, positioning fixture adjustment mark, and boat frame mark.

[0077] For example, Figure 3 This is a diagram illustrating various types of scratches on the surface of solar cells, such as... Figure 3 As shown, the standard images of scratch types, from left to right, are dry / wet basket card marks, belt marks, suction cup marks, and positioning fixture adjustment marks. Preferably, the process of drawing each of these scratch types is as follows: using computer-aided design software (preferably CAD), simulate and draw various possible mechanical process contact point graphics (drawn based on the contact points between the actual object and the battery cell, with an accuracy of 0.01mm; for example, drawing the outline of the scratch image that the belt might cause on the battery cell based on the point of contact between the belt and the battery cell). These graphics are stored in a library with a capacity of 1GB (i.e., establishing a mechanical process contact point library) for subsequent comparison and analysis.

[0078] Figure 4 It is based on Figure 1 The flowchart shown is a method for analyzing the proportion of scratches on battery cells. Optionally, the method further includes:

[0079] In step 401, the total number of battery cells produced within a preset time period and the number of battery cells for each type of scratch are obtained.

[0080] In step 402, the wear degree of the corresponding mechanical fixture within a preset time period is determined based on the proportion of the number of battery cells with each type of scratch to the total number of battery cells produced.

[0081] For example, the total number of solar cells produced within a preset time period and the number of solar cells with each type of scratch are obtained, and statistics are compiled on a daily / weekly / monthly / quarterly / yearly basis. The proportion of each scratch type of solar cell produced within this time period is calculated to determine the percentage of each scratch type. Furthermore, based on the proportion of each scratch type in the total number of solar cells, the wear level of the corresponding mechanical fixture within the preset time period can be determined, enabling precise maintenance and upkeep of the machinery.

[0082] The above steps can be automated through a computer program to determine the cause of scratches on battery cells, thereby improving the efficiency and accuracy of fault diagnosis and meeting the needs of large-scale production. Specifically, by writing a corresponding computer program, the above steps are automated. The computer program runs on Windows 10, with an Intel Core i7 processor, 8GB of memory, and a 500GB hard drive.

[0083] Figure 5 This is a structural block diagram of an image processing-based battery cell scratch type detection device according to an exemplary embodiment, such as... Figure 5 As shown, the device 500 includes:

[0084] The image acquisition module 510 acquires images of the battery cells to be inspected in real time during the battery cell production process.

[0085] The EL detection module 520 is connected to the image acquisition module 510. It uses a defect detection device to detect the battery cell under test and determine whether there are scratches on the surface of the battery cell under test.

[0086] The grayscale processing module 530 is connected to the EL detection module 520. If it is determined that there are scratches on the surface of the battery cell to be detected, the grayscale processing module 530 performs grayscale processing on the image to be detected to obtain the grayscale value of the image to be detected.

[0087] The contour restoration module 540 is connected to the grayscale processing module 530. It restores the shape contour of the scratched mechanical fixture on the image to be detected according to the grayscale value, and obtains the scratch contour image.

[0088] The scratch type determination module 550 is connected to the contour restoration module 540. Based on the scratch contour image and the preset scratch type standard image, the scratch type of the battery cell to be tested is determined by the correlation coefficient matching method.

[0089] Figure 6 It is based on Figure 5 The diagram shown is a structural block diagram of a scratch type determination module, such as... Figure 6 As shown, the scratch type determination module 550 includes:

[0090] The correlation coefficient calculation unit 551 determines the correlation coefficient ρ between the scratch contour image and the standard image for each scratch type through the correlation coefficient calculation formula.

[0091] The formula for calculating the correlation coefficient is: in,

[0092] g(x,y) represents the pixel values ​​of the standard scratch image, and g′(x,y) represents the pixel values ​​of the scratch contour image.

[0093] The scratch type determination unit 552 is connected to the correlation coefficient calculation unit 551. If the correlation coefficient between the scratch contour image and the scratch type standard image is greater than a preset coefficient threshold, the scratch type corresponding to the scratch type standard image is determined as the scratch type of the battery cell to be tested.

[0094] Figure 7 It is based on Figure 5 The diagram shown is a structural block diagram of a contour reconstruction module, such as... Figure 7 As shown, the contour reconstruction module 540 includes:

[0095] The scratch area determination unit 541 disassembles the image to be detected according to the gray level value and determines the area with a gray level value of 0% as the scratch area.

[0096] The scratch contour acquisition unit 542 is connected to the scratch area determination unit 541. Based on the contour of the scratch area, it reconstructs the shape contour of the scratched mechanical fixture on the image to be detected and acquires the scratch contour image.

[0097] In summary, this invention discloses a method and apparatus for detecting scratch types in battery cells based on image processing. The method includes: acquiring images of the battery cell to be inspected in real time during the battery cell production process; inspecting the battery cell to be inspected using a defect detection device to determine whether scratches exist on the surface of the battery cell; if scratches are found on the surface of the battery cell to be inspected, performing grayscale processing on the image to be inspected to obtain grayscale values; reconstructing the shape and contour of the mechanical fixture with the scratch on the image to be inspected based on the grayscale values ​​to obtain a scratch contour image; and determining the scratch type of the battery cell to be inspected using a correlation coefficient matching method based on the scratch contour image and a preset scratch type standard image. This method can determine whether the battery cell is scratched and the type of scratch by processing and analyzing the battery cell image, reducing subjective errors in the manual inspection process, improving inspection efficiency and accuracy, and significantly reducing labor costs.

[0098] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0099] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0100] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

Claims

1. A battery piece scratch type detection method based on image processing, characterized in that, The method comprises: During the production of the battery sheet, a to-be-detected image of a to-be-detected battery sheet is collected in real time; The to-be-detected battery sheet is detected by a defect detection device to determine whether there is a scratch on the surface of the to-be-detected battery sheet; If it is determined that there is a scratch on the surface of the to-be-detected battery sheet, the to-be-detected image is subjected to grayscale processing to obtain a grayscale degree value of the to-be-detected image; The mechanical clamp shape profile of the scratch is restored on the to-be-detected image according to the grayscale degree value to obtain a scratch profile image; The type of the scratch of the to-be-detected battery sheet is determined by a correlation coefficient matching method according to the scratch profile image and a preset scratch type standard image; The type of the scratch of the to-be-detected battery sheet is determined by a correlation coefficient matching method according to the scratch profile image and a preset scratch type standard image, which comprises: The correlation coefficient between the scratch contour image and the standard image of each scratch type is determined by a correlation coefficient calculation formula ; The correlation coefficient calculation formula is: wherein, , , is a pixel value of a scratch type standard image, is a pixel value of a scratch contour image; If the correlation coefficient between the scratch profile image and the scratch type standard image is greater than a preset coefficient threshold, the type of the scratch corresponding to the scratch type standard image is determined as the type of the scratch of the to-be-detected battery sheet; The method further comprises: The total number of battery sheets produced in a preset time period and the number of battery sheets of each scratch type are obtained; The wear degree of the corresponding mechanical clamp in the preset time period is determined according to the proportion of the number of battery sheets of each scratch type in the total number of battery sheets produced.

2. The image processing-based cell piece scratch type detection method according to claim 1, characterized in that, The defect detection device is an EL tester, and the to-be-detected image of the to-be-detected battery sheet is collected in real time, which comprises: An industrial camera arranged above the battery sheet collects images of the battery sheet at a frequency of 10 images per second; The image with the highest definition is selected as the to-be-detected image of the to-be-detected battery sheet.

3. The image processing-based cell piece scratch type detection method according to claim 1, characterized in that, If it is determined that there is a scratch on the surface of the to-be-detected battery sheet, the to-be-detected image is subjected to grayscale processing to obtain a grayscale degree value of the to-be-detected image, which comprises: If it is determined that there is a scratch on the surface of the to-be-detected battery sheet, the to-be-detected image is subjected to grayscale processing, the grayscale degree value is divided into 0-100% according to the grayscale degree value, and the grayscale degree value of the to-be-detected image is obtained.

4. The image processing-based cell piece scratch type detection method according to claim 3, characterized in that, The mechanical clamp shape profile of the scratch is restored on the to-be-detected image according to the grayscale degree value to obtain a scratch profile image, which comprises: The to-be-detected image is disassembled according to the grayscale degree value, and the area with a grayscale degree value of 0% is determined as the scratch area; The mechanical clamp shape profile of the scratch is restored on the to-be-detected image according to the profile of the scratch area to obtain a scratch profile image.

5. The image processing-based cell piece scratch type detection method according to claim 1, characterized in that, The method further comprises: Each of the scratch type standard images is drawn according to the contact point between the mechanical clamp and the battery sheet during the production of the battery sheet, wherein the types of the scratch include dry flower basket card point printing, wet flower basket card point printing, belt printing, suction cup printing, positioning clamp adjustment printing, and boat frame printing.

6. An image processing-based cell piece scratch type detection device, characterized by, The device comprises: An image collection module collects a to-be-detected image of a to-be-detected battery sheet in real time during the production of the battery sheet; An EL detection module connected to the image collection module detects the to-be-detected battery sheet by a defect detection device to determine whether there is a scratch on the surface of the to-be-detected battery sheet; The gray processing module is connected with the EL detection module, and if it is determined that the surface of the battery piece to be detected has a scratch, the gray processing is performed on the image to be detected to obtain a gray degree value of the image to be detected. The contour restoration module is connected with the gray processing module, and a mechanical clamp shape contour of the scratch is restored on the image to be detected according to the gray degree value to obtain a scratch contour image. The scratch type determination module is connected with the contour restoration module, and the scratch type of the battery piece to be detected is determined by a correlation coefficient matching method according to the scratch contour image and a preset scratch type standard image. The scratch type determination module comprises: The correlation coefficient calculation unit determines the correlation coefficient between the scratch contour image and each standard image of the scratch type by a correlation coefficient calculation formula ; The correlation coefficient calculation formula is: wherein, , , is a pixel value of a scratch type standard image, is a pixel value of a scratch contour image; The scratch type determination unit is connected with the correlation coefficient calculation unit, and if the correlation coefficient between the scratch contour image and the scratch type standard image is greater than a preset coefficient threshold, the scratch type corresponding to the scratch type standard image is determined as the scratch type of the battery piece to be detected. The total number of the battery pieces produced in a preset time period and the number of the battery pieces of each scratch type are obtained. According to the proportion of the number of the battery pieces of each scratch type in the total number of the battery pieces produced, the wear degree of the corresponding mechanical clamp in the preset time period is determined.

7. The image processing-based cell piece scratch type detection apparatus according to claim 6, characterized by, The contour restoration module comprises: The scratch area determination unit decomposes the image to be detected according to the gray degree value, and determines the area with a gray degree value of 0% as a scratch area. The scratch contour acquisition unit is connected with the scratch area determination unit, and restores the mechanical clamp shape contour of the scratch on the image to be detected according to the contour of the scratch area to obtain a scratch contour image.

Images