Image filtering method and device, storage medium and computer program product

Abstract

The invention discloses an image filtering method and device, a storage medium and a computer program product. The method comprises the following steps: identifying a pole piece area image where a battery cell pole piece is located from a battery cell cross-section image; dividing the pole piece region image into a plurality of sub-region images; obtaining the signal-to-noise ratio of each sub-region image; respectively determining a filtering parameter corresponding to the signal-to-noise ratio of each sub-region image from an incidence relation between a preset signal-to-noise ratio and the filtering parameter; and performing filtering processing on the corresponding sub-region image in the cell cross-section image by using the filtering parameter corresponding to each sub-region image to obtain a target cell cross-section image. According to the scheme provided by the invention, the quality of the pole piece area image can be improved, and the accuracy of cell defect detection is further improved.

Description

Technical Field

[0001] The present application relates to the field of defect detection technology, and in particular to an image filtering method, device, storage medium and computer program product. Background Art

[0002] In order to improve battery performance, it is necessary to carry out defect detection on the battery cells. If the distance between adjacent cathode plates and anode plates of the battery cells is too large, it is easy to cause lithium deposition in the battery, thereby reducing battery performance and shortening battery life.

[0003] In the related art, the distance between adjacent cathode and anode pieces of a battery cell can be detected by image detection. Therefore, improving the image quality can improve the detection accuracy of the distance between the cathode and anode pieces, which is of great significance for defect detection of battery cells. Summary of the invention

[0004] The present application provides an image filtering method, device, storage medium and computer program product, which can improve the quality of pole area images and thus improve the accuracy of battery cell defect detection.

[0005] In a first aspect, the present application provides an image filtering method, the method comprising: identifying a pole piece region image where a battery cell pole piece is located from a battery cell cross-sectional image; dividing the pole piece region image into a plurality of sub-region images; obtaining a signal-to-noise ratio of each sub-region image; determining the filtering parameters corresponding to the signal-to-noise ratio of each sub-region image from a correlation relationship between a preset signal-to-noise ratio and a filtering parameter; and filtering the corresponding sub-region images in the battery cell cross-sectional image using the filtering parameters corresponding to each sub-region image to obtain a target battery cell cross-sectional image.

[0006] In an embodiment of the present application, by calculating the signal-to-noise ratio of different sub-region images in the pole piece region image, the sub-region image is filtered using filtering parameters corresponding to the signal-to-noise ratio, thereby achieving adaptive multi-scale filtering of the pole piece region image, making the filtering of the pole piece region image more targeted, improving the quality of the pole piece region image, providing a basis for improving the detection accuracy of the distance between the cathode pole piece and the anode pole piece, and making the defect detection of the battery cell more accurate.

[0007] In some embodiments, identifying a pole piece region image where a battery cell pole piece is located from a battery cell cross-sectional image includes: filtering the battery cell cross-sectional image according to a first filtering threshold to obtain a filtered battery cell cross-sectional image, wherein the first filtering threshold is determined by a pixel grayscale value of the battery cell cross-sectional image; dilating the filtered battery cell cross-sectional image to obtain a region contour of at least one battery cell bending region; determining a target region contour from at least one region contour according to an arrangement position of the region contour in the battery cell cross-sectional image; and determining the battery cell bending region corresponding to the target region contour as the pole piece region image.

[0008] Adaptive filtering of the cell cross-sectional image using the first filtering threshold adapted to the cell cross-sectional image can improve the quality of the cell cross-sectional image and provide a basis for subsequent detection of the cell electrode spacing. In addition, by identifying the cell bending area in the cell cross-sectional image, the number of cells and their arrangement positions can be determined, so that the cells to be tested can be selected according to the test requirements, reducing the risk of false detection and missed detection of cells.

[0009] In some embodiments, before filtering the battery cell cross-section image according to a first filtering threshold to obtain the filtered battery cell cross-section image, image convolution is performed on the battery cell cross-section image through a first convolution kernel to obtain first contour data corresponding to the battery cell cross-section image; the pixel grayscale standard deviation corresponding to the first contour data is calculated; when the first coefficient is greater than a preset first standard deviation threshold, the first coefficient is determined as the first filtering threshold, wherein the first coefficient is the product of a preset first standard deviation coefficient and the pixel grayscale standard deviation; when the first coefficient is less than or equal to the first standard deviation threshold, the first standard deviation threshold is determined as the first filtering threshold.

[0010] Since the first filtering threshold adaptively changes according to the different cross-sectional images of the battery cell, that is, in the embodiment of the present application, a filtering threshold adapted to the cross-sectional image of the battery cell is used to filter the cross-sectional image of the battery cell, therefore, the above scheme can improve the accuracy of image filtering and provide a basis for accurate detection of the spacing between battery cell pole pieces.

[0011] In some embodiments, obtaining the signal-to-noise ratio of each sub-region image includes: for each sub-region image, respectively calculating the pixel variance of all pixels in the sub-region image; and determining the signal-to-noise ratio of the sub-region image according to the pixel variance.

[0012] By calculating the signal-to-noise ratio of each sub-region image, it is convenient to use different filtering parameters to denoise each sub-region image separately, realize accurate filtering of the polar region image, and improve the image quality.

[0013] In some embodiments, when the battery cell is a wound battery cell, the pole piece area image is divided into multiple sub-area images, including: obtaining the center point of the wound battery cell in the pole piece area image; rotating the pole piece area image at preset angle intervals with the center point as the rotation center to obtain pole piece area images corresponding to multiple rotation angles; determining a to-be-processed area from the pole piece area image corresponding to each rotation angle, wherein the to-be-processed area is an image area determined based on the center point and a preset number of pixels; and dividing the to-be-processed area into multiple sub-area images based on a window of a preset size.

[0014] By dividing the pole area image into multiple sub-area images, when filtering the pole area image, the filtering parameters corresponding to the sub-area images can be used to adaptively filter the corresponding sub-area images. Compared with using a single filtering parameter to filter the entire pole area image, the quality of the filtered pole area image can be improved.

[0015] In some embodiments, obtaining the center point of a wound battery cell in a pole piece area image includes: determining a bending direction of the wound battery cell based on multiple target positions corresponding to a bending area of ​​the battery cell in the pole piece area image, wherein the multiple target positions are located on a contour of the bending area of ​​the battery cell, and the bending direction is a direction pointed by a protrusion of the bending area of ​​the battery cell; determining the center point of the wound battery cell from the pole piece area image based on the bending direction.

[0016] The bending direction of the battery cell bending area is different, and the bending direction of the wound battery cell is also different. Therefore, the accuracy of determining the bending direction of the battery cell bending area can be improved according to multiple target positions on the battery cell bending area, and then the center point of the wound battery cell is determined according to the bending direction, which can improve the accuracy of confirming the center point of the wound battery cell.

[0017] In some embodiments, the bending direction of the wound battery cell is determined based on multiple target positions corresponding to the battery cell bending area in the pole piece area image, including: dividing the battery cell bending area in the pole piece area image into a first bending area and a second bending area along a first direction; obtaining a first coordinate value of the first target position of the first bending area in the first direction and a second coordinate value of the second target position of the first bending area in the first direction, wherein the distance between the first target position and the second target position in the first direction is the length of the first bending area in the first direction; obtaining a third coordinate value of the third target position of the second bending area in the first direction and a fourth coordinate value of the fourth target position of the second bending area in the first direction, wherein the distance between the third target position and the fourth target position in the first direction is the length of the second bending area in the first direction; determining the bending direction of the wound battery cell based on the first coordinate value, the second coordinate value, the third coordinate value and the fourth coordinate value.

[0018] The bending direction of the battery cell bending area is determined by the coordinate values ​​of the four target positions of the battery cell bending area, which facilitates the determination of the center point of the subsequent winding battery cell.

[0019] In some embodiments, the bending direction of the wound battery cell is determined based on the first coordinate value, the second coordinate value, the third coordinate value and the fourth coordinate value, including: when the second coordinate value is less than the first coordinate value and the fourth coordinate value is greater than the third coordinate value, the bending direction is determined as the second direction; when the second coordinate value is greater than the first coordinate value and the fourth coordinate value is less than the third coordinate value, the bending direction is determined as the third direction, wherein the second direction and the third direction are opposite directions to each other.

[0020] The bending direction of the wound battery cell is determined based on the coordinate value corresponding to the target position in the pole piece area image, and the center point of the wound battery cell is determined, so that the pole piece area image is rotated at multiple angles and divided into multiple sub-area images, and then each sub-area image is filtered in a targeted manner to improve the quality of the pole piece area image.

[0021] In some embodiments, the center point of the wound battery cell is determined from the pole piece area image based on the bending direction, including: filtering the pole piece area image according to a second filtering threshold to obtain a filtered pole piece area image, wherein the second filtering threshold is determined by the pixel grayscale value of the pole piece area image; extracting a target contour having a contour length greater than a preset length from the filtered pole piece area image; and determining the endpoint of the target contour in the bending direction as the center point.

[0022] By determining the center point of the wound battery cell, the pole piece area image can be rotated at multiple angles to facilitate the subsequent division of the sub-area image, and then each sub-area image can be filtered in a targeted manner to improve the quality of the pole piece area image.

[0023] In some embodiments, the corresponding sub-region images in the battery cell cross-section image are filtered using filtering parameters corresponding to each sub-region image to obtain a target battery cell cross-section image, including: filtering the corresponding sub-region images using filtering parameters corresponding to each sub-region image to obtain a plurality of filtered sub-region images; and splicing the plurality of filtered sub-region images according to the segmentation position corresponding to each sub-region image to obtain a target battery cell cross-section image.

[0024] The sub-region image is filtered using filtering parameters corresponding to the signal-to-noise ratio of the sub-region image, thereby realizing adaptive multi-scale filtering of the pole region image, making the filtering of the pole region image more targeted and improving the quality of the pole region image.

[0025] In some embodiments, after filtering the corresponding sub-region images in the battery cell cross-sectional image using filtering parameters corresponding to each sub-region image to obtain the target battery cell cross-sectional image, the cathode pole piece contour and the anode pole piece contour of the battery cell are identified from the target battery cell cross-sectional image; the distance between adjacent cathode pole piece contours and anode pole piece contours is calculated to obtain the pole piece distance; and the abnormal pole piece position whose pole piece distance is greater than the preset pole piece distance is identified from the target battery cell cross-sectional image.

[0026] By performing defect detection on the battery cell through the cross-sectional image of the target battery cell, the accuracy of battery cell defect detection can be improved.

[0027] In the second aspect, the present application also provides an image filtering device, which includes: an image recognition module, used to identify the pole piece area image where the battery cell pole piece is located from the battery cell cross-sectional image; an image division module, used to divide the pole piece area image into multiple sub-region images; a data acquisition module, used to obtain the signal-to-noise ratio of each sub-region image; a parameter determination module, used to determine the filtering parameters corresponding to the signal-to-noise ratio of each sub-region image from the correlation relationship between the preset signal-to-noise ratio and the filtering parameters; an image filtering module, used to use the filtering parameters corresponding to each sub-region image to filter the corresponding sub-region images in the battery cell cross-sectional image, so as to obtain the target battery cell cross-sectional image.

[0028] In a third aspect, the present application provides a readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the image filtering method as described in the first aspect is implemented.

[0029] In a fourth aspect, the present application provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the image filtering method as described in the first aspect.

[0030] 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

[0031] The features, advantages and technical effects of exemplary embodiments of the present application will be described below with reference to the accompanying drawings.

[0032] Figure 1 This is a flowchart of an image filtering method according to an embodiment of the present application; Figure 2 A schematic diagram of a cross-sectional image of a wound battery cell according to an embodiment of the present application; Figure 3A schematic diagram of an image of a pole piece region according to an embodiment of the present application; Figure 4 A schematic diagram of the regional outline of a battery cell bending region according to an embodiment of the present application; Figure 5 A schematic flow chart of a method for determining a first filtering threshold value according to an embodiment of the present application; Figure 6 This is a schematic diagram of an image of a pole piece region after rotation according to an embodiment of the present application; Figure 7 A schematic diagram of a regional outline of a battery cell bending region according to an embodiment of the present application; Figure 8 A schematic diagram of a pole piece region image after filtering according to an embodiment of the present application; Fig. 9 A schematic diagram of a center point area corresponding to a pole piece area image according to an embodiment of the present application; Fig.10 A schematic diagram of a target profile of an embodiment of the present application; Fig.11 A schematic diagram of a sub-region image after filtering according to an embodiment of the present application; Fig.12 A schematic diagram of a cross-sectional image of a target battery cell according to an embodiment of the present application; Fig.13 FIG. 1 is a schematic diagram of an image filtering device according to another embodiment of the present application.

[0033] In the drawings, the drawings are not necessarily drawn to scale. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0035] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" in the specification and claims of this application and the above-mentioned figure descriptions and any variations thereof are intended to cover non-exclusive inclusions.

[0036] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "multiple" is more than two, unless otherwise clearly and specifically defined.

[0037] Mentioning "embodiment" in the embodiments of the present application means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present application. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiment described in the embodiments of the present application can be combined with other embodiments.

[0038] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0039] In the description of the embodiments of the present application, the term "multiple" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0040] In the description of the embodiments of the present application, the technical terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, which are only for the convenience of describing the embodiments of the present application and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the embodiments of the present application.

[0041] In the description of the embodiments of the present application, unless otherwise clearly specified and limited, technical terms such as "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, and it can be the internal connection of two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0042] In the production process of lithium batteries, it is necessary to inspect the appearance and internal structure of the lithium battery cells. In the defect inspection of the cells, if the interval between the adjacent anode and cathode plates in the cell is greater than the specified value, the cell is prone to lithium deposition, and lithium deposition will cause the appearance of metallic lithium inside the battery, which will not only reduce the performance of the battery, but also limit the battery capacity, significantly shorten the battery life, and cause safety hazards for lithium batteries. In severe cases, it may even cause major safety accidents such as battery short circuit and explosion.

[0043] In the related art, a CT cross-sectional image of a large surface of a battery is obtained by a standard CT (Computed Tomography) device, and then the intervals between adjacent anode and cathode plates are manually calculated using the measurement function of the software provided on the CT device.

[0044] In this method, the accuracy of the test results obtained by manual calculation is low, and the test personnel for each test may be different, resulting in differences in the test results for each test. The consistency of the test results is poor, and it is impossible to measure strictly in the prescribed direction. There is a large degree of subjectivity in human measurement. Moreover, as the number of tests increases, the number of test personnel required also increases, which increases the cost of battery cell testing.

[0045] To solve the above problems, in the related art, the cell detection system can use the CT cross-sectional image of the cell to detect the cell pole piece spacing. However, during the CT imaging process, the image quality will be affected by many factors, such as dose, detector aperture, reconstruction range, scanning layer thickness and reconstruction algorithm. Under the multiple influences of these factors, the contrast of CT imaging is low, and in severe cases, some areas in the CT cross-sectional image will be blurred, which is not conducive to the subsequent detection of the cell pole pieces.

[0046] It can be seen that the image quality of the cell CT cross-sectional image will affect the detection results of the cell electrode spacing.

[0047] In the related art, the quality of CT cross-sectional images can be improved by image filtering. However, since the signal-to-noise ratios of different regions of CT cross-sectional images vary greatly, using the same filtering parameters to filter different image regions will make it impossible to filter out the noise in some regions, thus reducing the image quality of CT cross-sectional images.

[0048] In order to solve the problems existing in the related art, the embodiments of the present application provide an image filtering method, device, storage medium and computer program product. Among them, the method adopts a multi-scale filtering algorithm, determines different filtering parameters according to the signal-to-noise ratio of different areas in the CT cross-sectional image, and then uses the filtering parameters corresponding to the signal-to-noise ratio to filter the corresponding areas in the CT cross-sectional image to improve the quality of the CT cross-sectional image, which provides a basis for the subsequent accurate identification of the anode and cathode plates and the accurate calculation of the interval between the anode and cathode plates.

[0049] The following is an introduction to the image filtering method provided in the embodiments of the present application.

[0050] In one embodiment, Figure 1 The flowchart of the image filtering method is shown. The method can be applied to a battery cell detection system. The battery cell detection system may include an image acquisition device for acquiring a battery cell cross-sectional image and a host computer. The host computer may process (for example, filter) the battery cell cross-sectional image acquired by the image acquisition device, and identify the cathode and anode plates in the processed battery cell cross-sectional image, and then detect the interval between adjacent cathode and anode plates. Figure 1 As shown, the method may include the following steps S101 to S105: Step S101, identifying the electrode region image where the battery cell electrode is located from the battery cell cross-sectional image.

[0051] In step S101, the battery cell may be a wound battery cell or a laminated battery cell. The battery cell cross-sectional image may be a cross-sectional image of a large surface of the battery cell, for example, Figure 2 It is a cross-sectional image of a wound cell, such as Figure 2 As shown, the cross-sectional image of the battery cell includes not only the battery cell but also non-battery cell structures, such as the battery casing, etc. In order to improve the accuracy of battery cell detection, in the embodiment of the present application, the host computer removes the portion other than the battery cell from the cross-sectional image of the battery cell and extracts the electrode area image containing the battery cell electrode. Figure 2 The cell cross-section image shown in the figure can be used to identify the cell, and the Figure 3 The image of the pole piece region is shown in Figure 2. Figure 3 As shown, the first region image includes multiple contours, including the contour of the battery cell anode electrode piece, the contour of the battery cell cathode electrode piece, and the contour of the battery shell.

[0052] It should be noted that in step S101, the cross-sectional image of the battery cell may be a CT image. In this scenario, the image acquisition device may be a CT device, that is, the cross-sectional image of the battery cell may be obtained by scanning the battery through the CT device to achieve non-destructive testing of the battery cell. Figure 2As shown, in the case where the battery includes multiple battery cells, the battery cell cross-section image includes cross-sections of multiple battery cells. In this scenario, the host computer can perform image filtering on the battery cell cross-section image of each battery cell in turn according to a certain order, for example, the arrangement order of the battery cells in the battery cell cross-section image.

[0053] In addition, it should be noted that in step S101, the host computer may filter the cross-sectional image of the battery cell by means of adaptive threshold filtering to obtain a pole piece region image.

[0054] Step S102, dividing the pole region image into a plurality of sub-region images.

[0055] In step S102, the host computer may divide the pole region image into a plurality of sub-region images according to a preset image division rule. For example, the host computer may divide the pole region image into a plurality of sub-region images of equal size, or may divide the pole region image into a plurality of sub-region images of different sizes according to the distribution of pixel gray values ​​in the pole region image.

[0056] Step S103, obtaining the signal-to-noise ratio of each sub-region image.

[0057] In step S103, the signal-to-noise ratio of the image is the ratio of the signal to the noise in the image, which can be used to evaluate the image quality, wherein the higher the signal-to-noise ratio, the better the image quality. In the embodiment of the present application, the signal-to-noise ratio of each sub-region image can be represented by the pixel variance in the sub-region image, wherein the larger the pixel variance of the sub-region image, the larger the signal-to-noise ratio corresponding to the sub-region image, that is, there is a positive correlation between the pixel variance of the sub-region image and the signal-to-noise ratio.

[0058] Step S104: determining the filtering parameters corresponding to the signal-to-noise ratio of each sub-region image from the association relationship between the preset signal-to-noise ratio and the filtering parameters.

[0059] In step S104, the filtering parameters may include but are not limited to the convolution kernel size, the standard deviation coefficient corresponding to the convolution kernel, the minimum standard deviation threshold, etc., wherein the filtering threshold for threshold filtering of the sub-region image can be determined according to the filtering parameters, and then the filtering threshold is used to filter the sub-region image.

[0060] In the embodiment of the present application, the correlation between the signal-to-noise ratio and the filtering parameters can be determined by the tester based on the historical test data. In practical applications, the correlation between the signal-to-noise ratio and the filtering parameters, or the filtering parameters can be adaptively adjusted based on the real-time detection effect. For example, when the signal-to-noise ratio of the sub-region image is A1, the filtering parameter B1 is used to filter the sub-region image; but the filtering effect is not good, for example, all non-battery core parts cannot be filtered out, or some features of the battery core are filtered out; after filtering the sub-region image using the filtering parameter B2, all non-battery core parts are filtered out, or the battery core features are relatively intact, then the filtering parameter corresponding to the signal-to-noise ratio A1 can be updated from B1 to B2.

[0061] Step S105 , using the filtering parameters corresponding to each sub-region image to perform filtering processing on the corresponding sub-region images in the battery cell cross-sectional image, to obtain the target battery cell cross-sectional image.

[0062] After determining the filtering parameters corresponding to each sub-region image, the corresponding sub-region image is filtered using the corresponding filtering parameters instead of using the same filtering parameters to filter all sub-region images, so that the filtering of the pole region image is more targeted and the quality of the pole region image is improved.

[0063] Based on the scheme defined by the above steps S101 to S105, it can be known that in an embodiment of the present application, by calculating the signal-to-noise ratio of different sub-region images in the pole piece region image, the sub-region image is filtered using filtering parameters corresponding to the signal-to-noise ratio, thereby realizing adaptive multi-scale filtering of the pole piece region image, making the filtering of the pole piece region image more targeted, improving the quality of the pole piece region image, providing a basis for improving the detection accuracy of the distance between the cathode pole piece and the anode pole piece, and making the defect detection of the battery cell more accurate.

[0064] The following is an introduction to the implementation process of the method provided in the embodiment of the present application.

[0065] During the image processing, the host computer first pre-processes the cross-sectional image of the battery cell, such as filtering, dilation, and data extraction, so as to obtain the electrode area image where the battery cell electrode is located.

[0066] Specifically, the host computer filters the battery cell cross-section image according to a first filtering threshold to obtain a filtered battery cell cross-section image; then, the filtered battery cell cross-section image is expanded to obtain a regional contour of at least one battery cell bending region; then, according to the arrangement position of the regional contour in the battery cell cross-section image, a target region contour is determined from at least one region contour; finally, the battery cell bending region corresponding to the target region contour is determined as the pole piece region image.

[0067] In the above embodiment, the first filter threshold is determined by the pixel grayscale value of the cell cross-sectional image, that is, the first filter threshold is adaptively obtained according to the cell cross-sectional image, and is not a fixed value. For different cell cross-sectional images, the corresponding filter threshold is also different. Using the first filter threshold adapted to the cell cross-sectional image to filter the cell cross-sectional image can more accurately identify the area where the cell pole piece is located, providing a basis for improving image quality.

[0068] In one example, after obtaining the first filtering threshold, the host computer may determine the pixel grayscale mean value in each first convolution kernel in the cell cross-sectional image and the first filtering threshold as the target filtering threshold. For example, for each first convolution kernel, the host computer determines the sum of the pixel grayscale mean value corresponding to the first convolution kernel and the first filtering threshold as the target filtering threshold. Then, for each pixel point in the cell cross-sectional image, filtering is performed using formula (1): (1) In formula (1), is the gray value of a pixel in the first convolution kernel; is the grayscale mean of the pixels in the first convolution kernel; is the first filtering threshold.

[0069] After the cell cross-sectional image is filtered, the host computer performs expansion processing on the filtered cell cross-sectional image. In the embodiment of the present application, considering the thickness of the cathode electrode of the cell, the filtered cell cross-sectional image is expanded with a pixel unit of an expansion coefficient of 20*20. For example, Figure 2 After filtering the cell cross-section image shown in FIG. 1 , the filtered cell cross-section image is expanded to obtain Figure 4 The area outline of at least one cell bending area is shown. The number of the area outlines of the cell bending areas is the number of cells, and the arrangement positions of the area outlines of the cell bending areas in the cell cross-sectional image are the arrangement positions of the cells in the battery pack.

[0070] In the scenario where the battery pack includes multiple cells, the host computer can select the cells in a certain order. Figure 4 The target area contour to be processed is selected from the area contour of the cell bending area shown in the figure. The cell corresponding to the target area contour is the cell to be further processed this time. Figure 4 In the above example, the host computer selects the first area contour according to the battery cell detection sequence, and the host computer determines the area of ​​the battery cell corresponding to the first area contour in the battery cell cross-section image as the electrode area image.

[0071] Adaptive filtering of the cell cross-sectional image using the first filtering threshold adapted to the cell cross-sectional image can improve the quality of the cell cross-sectional image and provide a basis for subsequent detection of the cell electrode spacing. In addition, by identifying the cell bending area in the cell cross-sectional image, the number of cells and their arrangement positions can be determined, so that the cells to be tested can be selected according to the test requirements, reducing the risk of false detection and missed detection of cells.

[0072] In one embodiment, the first filtering threshold can be obtained by Figure 5 The method shown is to determine, Figure 5 As shown, the method for determining the first filtering threshold includes the following steps S501 to S504: Step S501, performing image convolution on the battery cell cross-section image using a first convolution kernel to obtain first contour data corresponding to the battery cell cross-section image; Step S502, calculating the pixel grayscale standard deviation corresponding to the first contour data; Step S503, when the first coefficient is greater than a preset first standard deviation threshold, determining the first coefficient as a first filtering threshold, wherein the first coefficient is a product of a preset first standard deviation coefficient and a pixel grayscale standard deviation; Step S504: when the first coefficient is less than or equal to the first standard deviation threshold, determine the first standard deviation threshold as the first filtering threshold.

[0073] In the above steps S501 to S504, the first convolution kernel can be a convolution kernel for performing image convolution along the first direction and the fourth direction, wherein the first direction is perpendicular to the fourth direction, the first direction can be the horizontal direction in the cross-sectional image of the battery cell, and the fourth direction can be the vertical direction in the cross-sectional image of the battery cell. As an example, the size of the first convolution kernel can be 301*11. In an embodiment of the present application, the host computer traverses all pixels in the cross-sectional image of the battery cell with the above-mentioned first convolution kernel, and calculates the pixel grayscale standard deviation stdValue of all pixels in each convolution kernel, and sets the first standard deviation coefficient StdDevScale to 0.75, the minimum standard deviation threshold (ie, the first standard deviation threshold) AbsThreshold to 1.0, and calculates the first filtering threshold by formula (2): (2) In formula (2), is the first filtering threshold, is the first standard deviation coefficient, is the pixel grayscale standard deviation, is the first standard deviation threshold; is the first coefficient.

[0074] In formula (2), the first standard deviation threshold and the first standard deviation coefficient can be set by the inspection personnel according to actual needs.

[0075] It should be noted that, through steps S501 to S504, it can be known that in the embodiment of the present application, the first filtering threshold is not fixed, but adaptively changes according to the different cross-sectional images of the battery cell. That is, in the embodiment of the present application, the cross-sectional image of the battery cell is filtered using a filtering threshold that is compatible with the cross-sectional image of the battery cell. This can improve the accuracy of image filtering and provide a basis for accurate detection of the spacing between battery cell pole pieces.

[0076] In one embodiment, after identifying the pole piece region image from the cell cross-sectional image, the host computer may divide the pole piece region image into a plurality of sub-region images, so as to specifically adopt filtering parameters adapted to each sub-region image for filtering.

[0077] When the battery cell is a wound battery cell, the host computer obtains the center point of the wound battery cell in the pole piece area image, and rotates the pole piece area image according to a preset angle interval with the center point as the rotation center to obtain pole piece area images corresponding to multiple rotation angles; then, the area to be processed is determined from the pole piece area image corresponding to each rotation angle, and the area to be processed is divided into multiple sub-area images based on a window of a preset size.

[0078] In the above embodiment, the center point of the wound battery cell may be the center point of the innermost electrode sheet of the wound battery cell. For example, the center point of the battery cell in the bending area facing the right side is Figure 3 The center point of the wound battery cell can be determined by adaptive filtering. The method for determining the center point of the wound battery cell is described below and will not be described here.

[0079] After determining the center point of the wound battery cell, the host computer uses the warpAffine function in the opencv image library to rotate the pole area image with the above center point as the rotation center and the preset angle interval (for example, 10°) as the rotation angle. For example, the host computer Figure 3 After rotating the pole area image shown in FIG. 40 counterclockwise, the following Figure 6 The rotated pole piece area image is shown in Figure 2. In practical applications, the host computer uses Figure 3 The position shown is 0°, and the pole region image is rotated at multiple angles to obtain pole region images corresponding to different rotation angles.

[0080] For the pole piece area image corresponding to each rotation angle, the area to be processed is determined from the pole piece area image, wherein the area to be processed is an image area determined based on the center point and a preset number of pixels. For example, for the pole piece area image corresponding to a certain rotation angle, the pole piece area image corresponding to the rotation angle is rotated so that it is parallel to the X direction, and the area determined by 50 pixels above and below the center point of the battery cell (i.e., the preset number of pixels) is the area to be processed.

[0081] After obtaining the area to be processed, the host computer may use a window of fixed size to divide the area to be processed into multiple sub-region images. For example, the host computer uses a window with a size of 50*100 pixels to divide the area to be processed.

[0082] By dividing the pole area image into multiple sub-area images, when filtering the pole area image, the filtering parameters corresponding to the sub-area images can be used to adaptively filter the corresponding sub-area images. Compared with using a single filtering parameter to filter the entire pole area image, the quality of the filtered pole area image can be improved.

[0083] It can be seen from the introduction of the above embodiments that, for a wound battery cell, in the process of dividing the electrode region image into a plurality of sub-region images, the center point of the wound battery cell needs to be determined.

[0084] Specifically, the host computer can determine the bending direction of the wound battery cell according to multiple target positions corresponding to the battery cell bending area in the pole piece area image, and then determine the center point of the wound battery cell from the pole piece area image based on the bending direction.

[0085] In the above embodiment, the multiple target positions corresponding to the bending area of ​​the battery cell are located on the contour of the bending area of ​​the battery cell, for example, Figure 7 In the schematic diagram of the area contour of the battery cell bending area shown, multiple target positions include the leftmost coordinate point leftUpPoint and the rightmost coordinate point rightUpPoint of the first part (for example, the upper half) of the area contour; the leftmost coordinate point leftDownPoint and the rightmost coordinate point rightDownPoint of the second part (for example, the lower half) of the area contour.

[0086] In the above embodiment, the bending direction of the battery cell bending region is the direction in which the protrusion of the battery cell bending region points. Figure 3 The bending direction corresponding to the bending area of ​​the battery cell is horizontally to the right.

[0087] It should be noted that the bending direction of the battery cell bending area is different, and the bending direction of the wound battery cell is also different. Therefore, the accuracy of determining the bending direction of the battery cell bending area can be improved based on multiple target positions on the battery cell bending area, and then the center point of the wound battery cell can be determined according to the bending direction, which can improve the accuracy of confirming the center point of the wound battery cell.

[0088] Regarding the bending direction of the wound battery cell, the upper computer can divide the battery cell bending area in the pole piece area image into a first bending area and a second bending area along a first direction; then, obtain the first coordinate value of the first target position of the first bending area in the first direction and the second coordinate value of the second target position of the first bending area in the first direction, and obtain the third coordinate value of the third target position of the second bending area in the first direction and the fourth coordinate value of the fourth target position of the second bending area in the first direction; then based on the first coordinate value, the second coordinate value, the third coordinate value and the fourth coordinate value, the bending direction of the wound battery cell can be determined.

[0089] In the above embodiment, the first direction may be the X direction of the pole region image. For example, the host computer may use the crop operator in the opencv image library to crop the pole region image along the X direction. Figure 4 Any one of the cell bending regions is divided into a first bending region and a second bending region, wherein the first bending region and the second bending region correspond to Figure 7 The bending areas of the left and right area contours. Figure 7 As shown, the first bending area includes a first target position leftUpPoint and a second target position rightUpPoint, and the distance between the first target position and the second target position in the first direction is the length of the first bending area in the first direction, that is, the first target position is the position of the leftmost pixel point of the first bending area, and the second target position is the position of the rightmost pixel point of the first bending area; the second bending area includes a third target position leftDownPoint and a fourth target position rightDownPoint, and the distance between the third target position and the fourth target position in the first direction is the length of the second bending area in the first direction, that is, the third target position is the position of the leftmost pixel point of the second bending area, and the fourth target position is the position of the rightmost pixel point of the second bending area.

[0090] It should be noted that the coordinate values ​​of the four target positions of the battery cell bending region can determine the bending direction of the battery cell bending region, which is convenient for determining the center point of the subsequent winding battery cell.

[0091] After the coordinate values ​​of the four target positions are obtained, the bending direction of the wound battery cell can be determined according to the coordinate values ​​of the four target positions.

[0092] In one embodiment, when the second coordinate value is less than the first coordinate value and the fourth coordinate value is greater than the third coordinate value, the host computer determines the bending direction as the second direction; when the second coordinate value is greater than the first coordinate value and the fourth coordinate value is less than the third coordinate value, the host computer determines the bending direction as the third direction.

[0093] In the above embodiment, the second direction and the third direction are opposite directions to each other. For example, the second direction is a horizontal rightward direction, and the third direction is a horizontal leftward direction.

[0094] As an example, the bending direction of the above-mentioned wound battery cell can be determined by formula (3): (3) In formula (3), is the first coordinate value, is the second coordinate value, is the third coordinate value, is the fourth coordinate value; is a second direction, for example, horizontally to the left; The third direction is, for example, horizontally to the right.

[0095] The bending direction of the wound battery cell is determined based on the coordinate value corresponding to the target position in the pole piece area image, and the center point of the wound battery cell is determined, so that the pole piece area image is rotated at multiple angles and divided into multiple sub-area images, and then each sub-area image is filtered in a targeted manner to improve the quality of the pole piece area image.

[0096] It can be seen from the above embodiments that after determining the bending direction of the pole piece region image, the host computer can determine the center point of the wound battery cell from the pole piece region image based on the bending direction. Specifically, the host computer can filter the pole piece region image according to the second filtering threshold to obtain a filtered pole piece region image; then, extract a target contour whose contour length is greater than a preset length from the filtered pole piece region image; and then determine the endpoint of the target contour in the bending direction as the center point.

[0097] In the above embodiment, the second filter threshold is determined by the pixel grayscale value of the pole region image, wherein the second filter threshold is determined in the same manner as the first filter threshold, and the only difference is the convolution kernel size and the pixel grayscale standard deviation. For the second filter threshold, the convolution kernel size corresponding to the second filter threshold is larger than the convolution kernel size corresponding to the first filter threshold. In the embodiment of the present application, the convolution kernel size corresponding to the second filter threshold can be 301*301, and the standard deviation coefficient StdDevScale is 0.3.

[0098] It should be noted that using a larger convolution kernel size to filter the pole area image can filter out most of the features of the pole area image and retain only the features related to the center point of the wound battery cell, which facilitates accurate and rapid determination of the center point of the wound battery cell.

[0099] In one example, the host computer uses the second filtering threshold to filter the pole area image and performs a negation operation to obtain the following Figure 8 Then, 50 pixels are taken upward and downward from the center position of the fourth direction (e.g., the Y direction of the image) of the filtered pole region image to determine the center point area where the center point of the wound battery cell is located, for example, Figure 8 The center point area corresponding to the filtered pole area image shown can be Fig. 9 Then, the host computer can use the findcontours function in the opencv image library to Fig. 9 The image shown in the figure is calculated, and the contour with the largest contour length is selected to obtain the following Fig.10 The target outline is shown. Figure 8 It can be seen that the bending direction of the wound battery cell is horizontal to the right, so Fig.10 The right end point of the target contour is determined as the center point of the wound battery cell.

[0100] By determining the center point of the wound battery cell, the pole piece area image can be rotated at multiple angles to facilitate the subsequent division of the sub-area image, and then each sub-area image can be filtered in a targeted manner to improve the quality of the pole piece area image.

[0101] This completes the introduction to the polar region image division.

[0102] After dividing the pole region image into a plurality of sub-region images, the host computer calculates the signal-to-noise ratio corresponding to each sub-region image, and then determines the filtering parameters corresponding to each sub-region image according to the signal-to-noise ratio corresponding to each sub-region image.

[0103] In one embodiment, for each sub-region image, the host computer calculates the pixel variance of all pixels in the sub-region image respectively, and then determines the signal-to-noise ratio of the sub-region image according to the pixel variance.

[0104] In one example, after the host computer divides the pole region image using a fixed window of 50*100, the image size corresponding to each sub-region image is 50*100. At this time, the host computer calculates the average value of all pixel values ​​in each sub-region image, and then calculates the variance of all pixel values ​​in the sub-region image. The variance is used to characterize the signal-to-noise ratio corresponding to the sub-region image, where the larger the variance, the higher the signal-to-noise ratio, and the smaller the variance, the lower the signal-to-noise ratio.

[0105] It should be noted that by calculating the signal-to-noise ratio of each sub-region image, it is convenient to use different filtering parameters to denoise each sub-region image respectively, thereby achieving accurate filtering of the polar region image and improving the image quality.

[0106] Furthermore, after calculating the signal-to-noise ratio of each sub-region image, the filtering parameters for filtering each sub-region image can be determined based on the correlation between the signal-to-noise ratio and the filtering parameters, so that the upper computer can use the filtering parameters corresponding to each sub-region image to filter the corresponding sub-region images in the battery cell cross-section image.

[0107] Specifically, the host computer uses the filtering parameters corresponding to each sub-region image to filter the corresponding sub-region images respectively to obtain multiple filtered sub-region images; then, the multiple filtered sub-region images are spliced ​​according to the segmentation position corresponding to each sub-region image to obtain the target battery cell cross-section image.

[0108] In one embodiment, the host computer filters the sub-region image in the pole piece region image every 5° to obtain each filtered sub-region image, such as Fig.11 Then, multiple filtered sub-region images are spliced ​​according to the rotation angle of the polar region image to obtain the following: Fig.12 The target cell cross-sectional image shown is the filtered cell cross-sectional image.

[0109] The sub-region image is filtered using filtering parameters corresponding to the signal-to-noise ratio of the sub-region image, thereby realizing adaptive multi-scale filtering of the pole region image, making the filtering of the pole region image more targeted and improving the quality of the pole region image.

[0110] So far, the filtering of the cross-sectional image of the battery cell is achieved by adopting the solution provided in the embodiment of the present application.

[0111] It should be noted that the above is introduced by taking a wound battery cell as an example. The solution provided in the embodiment of the present application can also be applied to other types of battery cells, for example, laminated battery cells. Among them, for laminated battery cells, it is not necessary to determine the bending area and center point of the battery cell. It is only necessary to divide the pole area image corresponding to the laminated battery cell into a plurality of sub-area images, and then determine the filtering parameters corresponding to the signal-to-noise ratio of each sub-area image from the correlation relationship between the preset signal-to-noise ratio and the filtering parameters; and then use the filtering parameters corresponding to each sub-area image to filter the corresponding sub-area images in the battery cell cross-sectional image to obtain the target battery cell cross-sectional image.

[0112] Furthermore, after filtering the battery cell cross-sectional image using the method provided in the embodiment of the present application, the host computer can also use the filtered target battery cell cross-sectional image to determine the electrode spacing between adjacent anode and cathode electrodes of the battery cell to perform defect detection on the battery cell.

[0113] In one embodiment, the host computer identifies the cathode pole piece contour and the anode pole piece contour of the battery cell from the target battery cell cross-sectional image; then, calculates the distance between adjacent cathode pole piece contours and anode pole piece contours to obtain the pole piece distance; and then identifies the abnormal pole piece position whose pole piece distance is greater than the preset pole piece distance from the target battery cell cross-sectional image.

[0114] In one example, the host computer can use a trained neural network model and / or image processing algorithm to identify the contours of the cathode and cathode electrodes from the cross-sectional image of the target battery cell. Then, the host computer obtains the ray diagram corresponding to the battery cell type, and obtains the intersection coordinates corresponding to the cathode and cathode electrodes by multiplying the ray diagram with the contour images corresponding to the cathode and cathode electrodes. The electrode distance between adjacent cathode and cathode electrodes can be determined based on the intersection coordinates corresponding to the cathode and cathode electrodes. The host computer can determine the location of the electrode abnormality based on the size of the electrode distance between adjacent cathode and cathode electrodes and the preset cathode distance. The host computer can also mark the location of the electrode abnormality in the cross-sectional image of the target battery cell to facilitate the inspection personnel to analyze the cause of the battery cell abnormality.

[0115] It should be noted that the accuracy of cell defect detection can be improved by performing cell defect detection on the target cell cross-sectional image.

[0116] This completes the explanation of the method provided in the embodiments of the present application.

[0117] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0118] In one embodiment, Fig.13 A schematic diagram of an image filtering device is shown. Fig.13It can be seen that the device 1300 includes: an image recognition module 1301 , an image segmentation module 1302 , a data acquisition module 1303 , a parameter determination module 1304 and an image filtering module 1305 .

[0119] An image recognition module 1301 is used to recognize an image of a pole region where a pole piece of a battery cell is located from a cross-sectional image of the battery cell; An image division module 1302 is used to divide the pole region image into a plurality of sub-region images; A data acquisition module 1303 is used to acquire the signal-to-noise ratio of each sub-region image; The parameter determination module 1304 is used to determine the filtering parameters corresponding to the signal-noise ratio of each sub-region image from the association relationship between the preset signal-noise ratio and the filtering parameters; The image filtering module 1305 is used to perform filtering processing on the corresponding sub-region images in the battery cell cross-sectional image respectively by using filtering parameters corresponding to each sub-region image to obtain a target battery cell cross-sectional image.

[0120] In one embodiment, the image recognition module includes: a first filtering module, an expansion module, a contour selection module and a first image determination module. The first filtering module is used to filter the cell cross-sectional image according to a first filtering threshold to obtain a filtered cell cross-sectional image, wherein the first filtering threshold is determined by the pixel grayscale value of the cell cross-sectional image; the expansion module is used to expand the filtered cell cross-sectional image to obtain the regional contour of at least one cell bending region; the contour selection module is used to determine the target region contour from at least one region contour according to the arrangement position of the region contour in the cell cross-sectional image; the first image determination module is used to determine the cell bending region corresponding to the target region contour as the pole piece region image.

[0121] In one embodiment, the image filtering device further includes: a convolution module, a first calculation module, a first threshold determination module, and a second threshold determination module. The convolution module is used to filter the cell cross-sectional image according to the first filtering threshold to obtain the filtered cell cross-sectional image, and then perform image convolution on the cell cross-sectional image through the first convolution kernel to obtain the first contour data corresponding to the cell cross-sectional image; the first calculation module is used to calculate the pixel grayscale standard deviation corresponding to the first contour data; the first threshold determination module is used to determine the first coefficient as the first filtering threshold when the first coefficient is greater than the preset first standard deviation threshold, wherein the first coefficient is the product of the preset first standard deviation coefficient and the pixel grayscale standard deviation; the second threshold determination module is used to determine the first standard deviation threshold as the first filtering threshold when the first coefficient is less than or equal to the first standard deviation threshold.

[0122] In one embodiment, the data acquisition module is specifically used to calculate the pixel variance of all pixels in each sub-region image respectively; and determine the signal-to-noise ratio of the sub-region image according to the pixel variance.

[0123] In one embodiment, the battery cell is a wound battery cell, and the image division module includes: a center point acquisition module, a rotation module, a second image determination module, and a first division module. The center point acquisition module is used to obtain the center point of the wound battery cell in the pole piece region image; the rotation module is used to rotate the pole piece region image according to a preset angle interval with the center point as the rotation center to obtain pole piece region images corresponding to multiple rotation angles; the second image determination module is used to determine the area to be processed from the pole piece region image corresponding to each rotation angle, wherein the area to be processed is an image area determined based on the center point and a preset number of pixels; the first division module is used to divide the area to be processed into multiple sub-region images based on a window of a preset size.

[0124] In one embodiment, the center point acquisition module includes: a first direction determination module and a center point determination module. The first direction determination module is used to determine the bending direction of the wound battery cell according to multiple target positions corresponding to the battery cell bending area in the pole piece area image, wherein the multiple target positions are located on the contour of the battery cell bending area, and the bending direction is the direction pointed by the protrusion of the battery cell bending area; the center point determination module is used to determine the center point of the wound battery cell from the pole piece area image based on the bending direction.

[0125] In one embodiment, the first direction determination module includes: a cell segmentation module, a first coordinate acquisition module, a second coordinate acquisition module and a second direction determination module. The cell segmentation module is used to segment the cell bending area in the pole piece area image into a first bending area and a second bending area along the first direction; the first coordinate acquisition module is used to obtain the first coordinate value of the first target position of the first bending area in the first direction and the second coordinate value of the second target position of the first bending area in the first direction, wherein the distance between the first target position and the second target position in the first direction is the length of the first bending area in the first direction; the second coordinate acquisition module is used to obtain the third coordinate value of the third target position of the second bending area in the first direction and the fourth coordinate value of the fourth target position of the second bending area in the first direction, wherein the distance between the third target position and the fourth target position in the first direction is the length of the second bending area in the first direction; the second direction determination module is used to determine the bending direction of the wound cell based on the first coordinate value, the second coordinate value, the third coordinate value and the fourth coordinate value.

[0126] In one embodiment, the second direction determination module is specifically used to determine the bending direction as the second direction when the second coordinate value is less than the first coordinate value and the fourth coordinate value is greater than the third coordinate value; and to determine the bending direction as the third direction when the second coordinate value is greater than the first coordinate value and the fourth coordinate value is less than the third coordinate value, wherein the second direction and the third direction are opposite directions to each other.

[0127] In one embodiment, the center point determination module is specifically used to filter the pole area image according to a second filtering threshold to obtain a filtered pole area image, wherein the second filtering threshold is determined by the pixel grayscale value of the pole area image; extract a target contour having a contour length greater than a preset length from the filtered pole area image; and determine the endpoint of the target contour in the bending direction as the center point.

[0128] In one embodiment, the image filtering module is specifically used to filter the corresponding sub-region images respectively using the filtering parameters corresponding to each sub-region image to obtain multiple filtered sub-region images; and splice the multiple filtered sub-region images according to the segmentation position corresponding to each sub-region image to obtain the target battery cell cross-section image.

[0129] In one embodiment, the image filtering device also includes: a defect detection module, which is used to filter the corresponding sub-region images in the battery cell cross-sectional image using filtering parameters corresponding to each sub-region image, and after obtaining the target battery cell cross-sectional image, identify the cathode pole piece contour and the anode pole piece contour of the battery cell from the target battery cell cross-sectional image; calculate the distance between adjacent cathode pole piece contours and anode pole piece contours to obtain the pole piece distance; and identify the abnormal pole piece position whose pole piece distance is greater than the preset pole piece distance from the target battery cell cross-sectional image.

[0130] In one embodiment, the present application further provides a readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the above-mentioned image filtering method is implemented.

[0131] In one embodiment, the present application further provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the above-mentioned image filtering method.

[0132] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0133] The functional blocks shown in the above structural block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application-specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier. "Machine-readable medium" may include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROMs, flash memory, erasable ROMs (EROMs), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0134] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0135] The above describes various aspects of the present application with reference to the flowcharts and / or block diagrams of the image filtering method, device, storage medium and computer program product according to the embodiments of the present application. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the functions / actions specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and they should all be included in the scope of the claims and specification of the present application. In particular, as long as there is no structural conflict, the various technical features mentioned in the various embodiments can be combined in any way. The present application is not limited to the specific embodiments disclosed herein, but includes all technical solutions that fall within the scope of the claims.

Claims

1. An image filtering method, characterized in that: include: Identify the electrode area image where the battery cell electrode is located from the battery cell cross-sectional image; Dividing the pole piece region image into a plurality of sub-region images; Acquire the signal-to-noise ratio of each sub-region image; Determining the filtering parameters corresponding to the signal-to-noise ratio of each sub-region image from the association relationship between the preset signal-to-noise ratio and the filtering parameters; The corresponding sub-region images in the battery cell cross-sectional image are respectively filtered using the filtering parameters corresponding to each of the sub-region images to obtain a target battery cell cross-sectional image.

2. The method according to claim 1, characterized in that The step of identifying the electrode region image where the electrode of the battery cell is located from the cross-sectional image of the battery cell includes: Performing filtering processing on the battery cell cross-sectional image according to a first filtering threshold value to obtain a filtered battery cell cross-sectional image, wherein the first filtering threshold value is determined by a pixel grayscale value of the battery cell cross-sectional image; Performing expansion processing on the filtered cross-sectional image of the battery cell to obtain a regional outline of at least one bending region of the battery cell; Determining a target region contour from at least one of the region contours according to an arrangement position of the region contours in the battery cell cross-sectional image; The cell bending area corresponding to the target area contour is determined as the pole piece area image.

3. The method according to claim 2, characterized in that Before filtering the battery cell cross-sectional image according to the first filtering threshold to obtain the filtered battery cell cross-sectional image, the method further includes: Performing image convolution on the battery core cross-sectional image through a first convolution kernel to obtain first contour data corresponding to the battery core cross-sectional image; Calculating the pixel grayscale standard deviation corresponding to the first contour data; When the first coefficient is greater than a preset first standard deviation threshold, determining the first coefficient as the first filtering threshold, wherein the first coefficient is the product of the preset first standard deviation coefficient and the pixel grayscale standard deviation; When the first coefficient is less than or equal to the first standard deviation threshold, the first standard deviation threshold is determined as the first filtering threshold.

4. The method according to claim 1, characterized in that: The acquiring the signal-to-noise ratio of each sub-region image comprises: For each of the sub-region images, respectively calculating the pixel variance of all pixels in the sub-region image; A signal-to-noise ratio of the sub-region image is determined according to the pixel variance.

5. The method according to claim 1, characterized in that The battery cell is a wound battery cell, and the electrode area image is divided into a plurality of sub-area images, including: Obtaining the center point of the wound battery cell in the pole piece area image; Taking the central point as the rotation center, the pole piece region image is rotated at preset angle intervals to obtain pole piece region images corresponding to a plurality of rotation angles; Determining a region to be processed from the pole region image corresponding to each rotation angle, wherein the region to be processed is an image region determined based on the center point and a preset number of pixels; The area to be processed is divided into a plurality of sub-area images based on a window of a preset size.

6. The method according to claim 5, characterized in that The obtaining of the center point of the wound battery cell in the pole piece region image includes: Determine the bending direction of the wound battery cell according to a plurality of target positions corresponding to the battery cell bending region in the electrode region image, wherein the plurality of target positions are located on the contour of the battery cell bending region, and the bending direction is the direction pointed by the protrusion of the battery cell bending region; The center point of the wound battery cell is determined from the pole piece area image based on the bending direction.

7. The method according to claim 6, characterized in that The step of determining the bending direction of the wound battery cell according to a plurality of target positions corresponding to the battery cell bending region in the electrode region image comprises: Dividing the cell bending region in the pole piece region image into a first bending region and a second bending region along a first direction; Obtaining a first coordinate value of a first target position of the first bending region in the first direction and a second coordinate value of a second target position of the first bending region in the first direction, wherein the distance between the first target position and the second target position in the first direction is the length of the first bending region in the first direction; Obtaining a third coordinate value of a third target position of the second bending area in the first direction and a fourth coordinate value of a fourth target position of the second bending area in the first direction, wherein the distance between the third target position and the fourth target position in the first direction is the length of the second bending area in the first direction; The bending direction of the wound battery cell is determined based on the first coordinate value, the second coordinate value, the third coordinate value, and the fourth coordinate value.

8. The method according to claim 7, characterized in that The determining the bending direction of the wound battery cell based on the first coordinate value, the second coordinate value, the third coordinate value, and the fourth coordinate value includes: When the second coordinate value is smaller than the first coordinate value and the fourth coordinate value is larger than the third coordinate value, the bending direction is determined as the second direction; When the second coordinate value is greater than the first coordinate value and the fourth coordinate value is less than the third coordinate value, the bending direction is determined as a third direction, wherein the second direction and the third direction are opposite directions to each other.

9. The method according to any one of claims 6 to 8, characterized in that The determining the center point of the wound battery cell from the pole piece region image based on the bending direction includes: Filtering the pole region image according to a second filtering threshold to obtain a filtered pole region image, wherein the second filtering threshold is determined by a pixel grayscale value of the pole region image; Extracting a target contour whose contour length is greater than a preset length from the filtered pole piece region image; An endpoint of the target contour in the bending direction is determined as the center point.

10. The method according to claim 1, characterized in that The filtering process is performed on the corresponding sub-region images in the cell cross-section image by using the filtering parameters corresponding to each of the sub-region images to obtain the target cell cross-section image, including: Using the filtering parameters corresponding to each of the sub-region images, filtering the corresponding sub-region images respectively to obtain a plurality of filtered sub-region images; The plurality of filtered sub-region images are spliced ​​according to the segmentation position corresponding to each of the sub-region images to obtain the target battery cell cross-sectional image.

11. The method according to claim 1, characterized in that: After filtering the corresponding sub-region images in the cell cross-sectional image using the filtering parameters corresponding to each of the sub-region images to obtain the target cell cross-sectional image, the method further includes: Identify the cathode electrode piece contour and the anode electrode piece contour of the battery cell from the target battery cell cross-sectional image; Calculate the distance between the adjacent cathode pole piece contours and the anode pole piece contours to obtain the pole piece distance; An abnormal pole piece position whose pole piece distance is greater than a preset pole piece distance is identified from the target battery cell cross-sectional image.

12. An image filtering device, characterized in that: include: An image recognition module, used to recognize an image of a pole region where a pole piece of a battery cell is located from a cross-sectional image of the battery cell; An image division module, used for dividing the pole region image into a plurality of sub-region images; A data acquisition module, used for acquiring the signal-to-noise ratio of each sub-region image; A parameter determination module, used to determine the filtering parameters corresponding to the signal-to-noise ratio of each sub-region image from the association relationship between the preset signal-to-noise ratio and the filtering parameters; The image filtering module is used to filter the corresponding sub-region images in the battery cell cross-sectional image by using the filtering parameters corresponding to each of the sub-region images to obtain a target battery cell cross-sectional image.

13. A readable storage medium, characterized in that: Computer program instructions are stored on the readable storage medium, and when the computer program instructions are executed by the processor, the image filtering method according to any one of claims 1 to 11 is implemented.

14. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the image filtering method as described in any one of claims 1 to 11.

Images