Battery cell detection method and apparatus, electronic device, and storage medium

Abstract

Embodiments of the present application relate to the technical field of batteries. Disclosed are a battery cell detection method and apparatus, an electronic device, and a storage medium. The battery cell detection method comprises: extracting a battery cell area from a battery cell image obtained by means of ray imaging; and on the basis of the distribution of grayscale values of pixels in the battery cell area, obtaining a feeding position detection result of a battery cell. The battery cell detection method provided by the embodiments of the present application can improve the accuracy and credibility of battery cell detection results.

Description

Battery cell detection method and device, electronic device, and storage medium

[0001] Cross-reference to Related Applications

[0002] This application claims priority to Chinese Patent Application No. 202410518413.2, filed April 28, 2024, entitled “Battery cell detection method and device, electronic device, and storage medium,” the entire contents of which are incorporated herein by reference. TECHNICAL FIELD

[0003] The present application relates to the technical field of batteries, and specifically relates to a battery cell detection method and device, an electronic device, and a storage medium. BACKGROUND

[0004] In the structure of a battery, a battery cell is an important component. In order to improve the safety of the battery, the internal structure of the battery cell usually needs to be detected.

[0005] In the related art, the detection method for the internal structure of the battery cell is to detect the battery cell during the winding process of the battery cell, so as to determine whether the feeding position of the finished battery cell has defects. However, this method needs to be detected during the winding process of the battery cell, and if it is not disassembled and verified, it cannot be determined whether the detection result is applicable to the battery cell after being made into a finished product, which affects the accuracy and reliability of the battery cell detection result.

[0006] SUMMARY

[0007] In view of the above problems, the present application provides a battery cell detection method and device, an electronic device, and a storage medium, which can improve the accuracy and reliability of the battery cell detection result.

[0008] In a first aspect, an embodiment of the present application provides a battery cell detection method, which comprises: extracting a battery cell region from a battery cell image obtained by radiographic imaging; and obtaining a feeding position detection result of the battery cell according to the gray value distribution of pixels in the battery cell region.

[0009] In the technical solution of the embodiments of the present application, the cell region is extracted from the cell image obtained by using the ray imaging, so as to determine the feeding position detection result of the cell according to the gray scale distribution of the pixels in the cell region. Since the cathode feeding position overlaps with the anode feeding position, the number of layers of the cell in the overlapping part increases, resulting in that more layers need to be penetrated during the ray imaging, and the pixel gray scale value of the overlapping part in the cell region is low at this time. If the cathode feeding position does not overlap with the anode feeding position, the number of layers of the cell between the cathode feeding position and the anode feeding position is small, resulting in that fewer layers need to be penetrated during the ray imaging, and the pixel gray scale value between the cathode feeding position and the anode feeding position is high at this time. Therefore, by detecting the gray scale value distribution of the pixels in the cell region, whether the cathode feeding position and the anode feeding position of the finished cell overlap can be determined, so as to improve the accuracy and reliability of the cell detection result.

[0010] In some embodiments, the cell region is extracted from the cell image obtained by using the ray imaging, including: performing binaryzation processing on the cell image according to the gray scale value of each pixel in the cell image, to obtain a binaryzation cell image of the cell image; and obtaining the cell region in the cell image according to the connected region corresponding to the cell in the binaryzation cell image. Thus, the noise of the cell image can be effectively removed by using the binaryzation method, and the cell region can be accurately extracted from the cell image by using the connected region in the binaryzation image, thereby improving the accuracy of cell region identification.

[0011] In some embodiments, the binaryzation processing on the cell image according to the gray scale value of each pixel in the cell image, to obtain a binaryzation cell image of the cell image, includes: performing binaryzation processing on the cell image according to the gray scale value of each pixel in the cell image, to obtain an initial binaryzation image of the cell image; and sequentially performing image erosion and image dilation on the initial binaryzation image, to obtain the binaryzation cell image of the cell image. Thus, the binaryzation image of the cell image can be effectively denoised, the clarity and readability of the obtained binaryzation image are improved, and the accuracy of the cell region extracted by using the binaryzation image is further improved.

[0012] In some embodiments, the cell region in the cell image is obtained according to the connected region corresponding to the cell in the binaryzation cell image, including: obtaining a target region in the cell image according to the connected region; and performing image enhancement processing on the target region to obtain the cell region; wherein the image enhancement processing includes filtering processing on each column of pixels in the target region in the vertical direction of the feeding position, to highlight or enhance the features in the vertical direction, so that the features of the cathode feeding position and the anode feeding position in the obtained cell region are more highlighted, and the cathode feeding position and the anode feeding position can be more clearly observed, facilitating subsequent cell detection.

[0013] In some embodiments, the feeding position detection result of the battery cell is obtained according to the gray value distribution of the pixels in the battery cell region, including: obtaining the feeding position detection result of the battery cell according to the gray values of the pixels in each column in the vertical direction of the feeding position in the battery cell region. Thus, when detecting the battery cell, the feeding position feature of the battery cell can be effectively utilized for detection, and the accuracy of the feeding position detection result of the battery cell is improved.

[0014] In some embodiments, the feeding position detection result of the battery cell is obtained according to the gray values of the pixels in each column in the vertical direction of the feeding position in the battery cell region, including: comparing the gray values of the pixels in each column with a preset gray value to obtain the feeding position detection result of the battery cell; and the preset gray value is the minimum gray value detected from the battery cell region of the abnormal battery cell overlapping the feeding position. In this way, by comparing the gray values of the pixels in each column with the preset gray value, it can be quickly judged whether the feeding position of the battery cell overlaps, and the detection efficiency of the battery cell is improved.

[0015] In some embodiments, the feeding position detection result of the battery cell is obtained according to the gray values of the pixels in each column in the vertical direction of the feeding position in the battery cell region, including: obtaining the gray values of the pixels in each column in the vertical direction of the feeding position in the battery cell region; obtaining the gray value change data of the battery cell region according to the gray values of the adjacent pixels in each column; and obtaining the feeding position detection result of the battery cell according to the gray value change data. Thus, when judging the feeding position detection result of the battery cell, the overall distribution of the pixels in each column in the vertical direction of the feeding position is considered, and the accuracy of the feeding position detection result is improved.

[0016] In some embodiments, the feeding position detection result of the battery cell is obtained according to the gray value change data, including: determining the anode feeding position and the cathode feeding position of the battery cell according to the gray difference between the adjacent gray values in the gray value change data; and obtaining the overlapping distance between the anode feeding position and the cathode feeding position of the battery cell according to the anode feeding position and the cathode feeding position. Thus, the obtained feeding position detection result is more accurate, and subsequent adjustment of the battery cell is facilitated.

[0017] In some embodiments, the feeding position detection result of the battery cell is obtained according to the gray value change data, including: using a pre-trained preset model to detect the gray value change data to obtain the overlapping distance between the anode feeding position and the cathode feeding position of the battery cell; and the preset model is trained by the gray value change data corresponding to each battery cell region sample, and each battery cell region sample includes the battery cell region of a normal battery cell and the battery cell region of an abnormal battery cell.

[0018] In a second aspect, the present application provides a battery cell detection device, including:

[0019] The region extraction module is configured to extract a battery cell region from a battery cell image obtained by ray imaging; and the battery cell detection module is configured to obtain a feeding position detection result of the battery cell according to a gray value distribution of pixels in the battery cell region.

[0020] In the technical scheme of the embodiments, the battery cell region is extracted from the battery cell image obtained by ray imaging, and the feeding position detection result of the battery cell is determined according to the gray value distribution of the pixels in the battery cell region, thereby improving the accuracy and reliability of the battery cell detection result.

[0021] In some embodiments, the region extraction module is specifically configured to: perform binaryzation processing on the battery cell image according to the gray values of the pixels in the battery cell image, to obtain a binaryzation battery cell image of the battery cell image; and obtain the battery cell region in the battery cell image according to a connected region corresponding to the battery cell in the binaryzation battery cell image.

[0022] In some embodiments, the region extraction module is specifically configured to: perform binaryzation processing on the battery cell image according to the gray values of the pixels in the battery cell image, to obtain an initial binaryzation image of the battery cell image; and sequentially perform image erosion and image dilation on the initial binaryzation image, to obtain a binaryzation battery cell image of the battery cell image.

[0023] In some embodiments, the region extraction module is specifically configured to: obtain a target region in the battery cell image according to the connected region; and perform image enhancement processing on the target region, to obtain the battery cell region; wherein the image enhancement processing comprises performing filtering processing on each column of pixels in the target region and located in the vertical direction of the feeding position.

[0024] In some embodiments, the battery cell detection module is specifically configured to: obtain the feeding position detection result of the battery cell according to the gray values of each column of pixels in the battery cell region and located in the vertical direction of the feeding position.

[0025] In some embodiments, the battery cell detection module is specifically configured to: compare the gray values of the columns of pixels with a preset gray value, to obtain the feeding position detection result of the battery cell; wherein the preset gray value is a minimum gray value detected from a battery cell region of an abnormal battery cell overlapping the feeding position.

[0026] In some embodiments, the battery cell detection module is specifically configured to: obtain the gray values of each column of pixels in the battery cell region and located in the vertical direction of the feeding position; obtain gray value change data of the battery cell region according to the gray values of adjacent columns of pixels; and obtain the feeding position detection result of the battery cell according to the gray value change data.

[0027] In some embodiments, the battery cell detection module is specifically configured to: determine the anode charging position and the cathode charging position of the battery cell according to the gray value difference of adjacent gray values in the gray change data; and obtain the overlap distance between the anode charging position and the cathode charging position of the battery cell according to the anode charging position and the cathode charging position.

[0028] In some embodiments, the battery cell detection module is specifically configured to: use a pre-trained preset model to detect the gray change data to obtain the overlap distance between the anode charging position and the cathode charging position of the battery cell; and wherein the preset model is trained by the gray change data corresponding to each battery cell region sample, and each battery cell region sample includes a battery cell region of a normal battery cell and a battery cell region of an abnormal battery cell.

[0029] In a third aspect, the present application provides an electronic device, comprising a memory and a processor, the memory stores a computer program, and the processor executes the computer program to execute the method in the embodiments of the first aspect.

[0030] In a fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to execute the method in the embodiments of the first aspect.

[0031] In a fifth aspect, the present application provides a computer program product, which, when running on a computer, causes the computer to execute the method in the first aspect, any optional implementation manner of the first aspect. BRIEF DESCRIPTION OF DRAWINGS

[0032] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are intended to depict only preferred embodiments of the application, and therefore should not be considered to narrow the scope of the present application. Rather, the entire disclosure including the full description and the drawings are to be considered to define the scope of the application. In the drawings:

[0033] FIG. 1 is a first flowchart of a battery cell detection method according to some embodiments of the present application;

[0034] FIG. 2 is a schematic diagram of a battery cell image according to some embodiments of the present application;

[0035] FIG. 3 is a schematic diagram of a battery cell region according to some embodiments of the present application;

[0036] FIG. 4a is a schematic diagram of a battery cell structure with normal charging positions according to some embodiments of the present application;

[0037] FIG. 4b is a schematic diagram of a battery cell structure with abnormal charging positions according to some embodiments of the present application;

[0038] FIG. 5a is a schematic diagram of pixel distribution of a cell area of a normal cell according to some embodiments of the present application;

[0039] FIG. 5b is a schematic diagram of pixel distribution of a cell area of an abnormal cell according to some embodiments of the present application;

[0040] FIG. 6 is a second flowchart of a cell detection method according to some embodiments of the present application;

[0041] FIG. 7 is a schematic diagram of gray scale change data of a normal cell according to some embodiments of the present application;

[0042] FIG. 8 is a schematic diagram of gray scale change data of an abnormal cell according to some embodiments of the present application;

[0043] FIG. 9 is a third flowchart of a cell detection method according to some embodiments of the present application;

[0044] FIG. 10 is a schematic diagram of a structure of a cell detection device according to some embodiments of the present application;

[0045] FIG. 11 is a schematic diagram of a structure of an electronic device according to some embodiments of the present application.

[0046] The reference signs in the detailed description of the embodiments are as follows: 401 - area extraction module; 402 - cell detection module; 5 - electronic device; 501 - processor; 502 - memory; 503 - communication bus. DETAILED DESCRIPTION

[0047] The embodiments of the technical solutions of the present application will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and therefore only serve as examples, and cannot limit the protection scope of the present application.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs; the terms used herein are only for the purpose of describing specific embodiments of the present application, and are not intended to limit the present application; the terms "include" and "have" and any variations thereof in the specification and claims of the present application and the above description of drawings are intended to cover non-exclusive inclusion.

[0049] 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 "a plurality of" is two or more, unless otherwise explicitly specified.

[0050] Reference to“an embodiment” herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase“in an embodiment” in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily all referring to a common set of embodiments, although they can. Those skilled in the art will appreciate that the embodiments described herein can be combined with other embodiments in various ways.

[0051] In the description of the embodiments of the application, the term“and / or” only means an association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character“ / ” herein generally represents an“or” relationship between the associated objects before and after it.

[0052] In the description of the embodiments of the application, the term“a plurality of” means two or more (including two), and similarly, “a plurality of groups” means two or more groups (including two groups), and “a plurality of pieces” means two or more pieces (including two pieces).

[0053] In the structure of the battery, the battery cell is an important component. In the production process of the battery cell, due to the influence of the environment, process or other accidental factors, the internal structure of the produced battery cell may have the defect of overlapping of the anode and cathode material loading positions. Therefore, in order to improve the yield of the battery cell, the internal structure of the battery cell needs to be detected.

[0054] At present, the detection of the internal structure of the battery cell is to disassemble the battery cell to detect whether the anode and cathode material loading positions of the battery cell have defects. However, this method will cause damage to the battery cell and has poor safety. Therefore, in the related art, the battery cell can be detected by obtaining the winding image of the battery cell during the winding process of the battery cell to determine whether the material loading position of the finished product has defects. However, this method needs to be detected during the winding process of the battery cell, which affects the production efficiency of the battery cell, and after the battery cell is made into a finished product, if it is not disassembled and verified, it cannot be determined whether the detection result is applicable to the battery cell after the finished product is made. At the same time, due to the influence of the environment, process or other accidental factors, the final finished product battery cell may have differences in structure from the theoretical finished product battery cell made by using the winding process, which affects the accuracy and reliability of the battery cell detection result.

[0055] To solve the above technical problems, the embodiments of the present application extract the battery cell region from the battery cell image obtained by using the ray imaging, to determine the charging level detection result of the battery cell according to the gray scale distribution of the pixels in the battery cell region. Since the cathode charging level overlaps with the anode charging level, the number of layers of the battery cell in the overlapping part increases, resulting in more layers needing to be penetrated during the ray imaging. At this time, the pixel gray scale value of the overlapping part in the battery cell region is low. If the cathode charging level does not overlap with the anode charging level, the number of layers of the battery cell between the cathode charging level and the anode charging level is small, resulting in fewer layers needing to be penetrated during the ray imaging. At this time, the pixel gray scale value between the cathode charging level and the anode charging level is high. Therefore, by detecting the gray scale value distribution of the pixels in the battery cell region, it can be determined whether the cathode charging level and the anode charging level of the finished battery cell overlap, thereby improving the accuracy and reliability of the battery cell detection result.

[0056] The battery cell detection method, device, electronic equipment and storage medium disclosed by the embodiments of the present application can be applied to a server, for detecting whether the cathode and anode charging levels of the finished battery cell overlap. The server can be an independent server or a server cluster composed of multiple servers, and can also be a cloud server providing basic cloud computing services such as cloud service, cloud database, cloud computing, cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN, and big data and artificial intelligence sampling point equipment.

[0057] According to some embodiments of the present application, a battery cell detection method is provided, which can be applied to the aforementioned server. As shown in FIG. 1, the battery cell detection method includes:

[0058] S101, extracting a battery cell region from a battery cell image obtained by ray imaging;

[0059] S102, obtaining a charging level detection result of the battery cell according to the gray scale value distribution of the pixels in the battery cell region.

[0060] The ray imaging refers to an imaging technology of using a ray beam to pass through a measured object to obtain a projection of the measured object, such as X-Ray imaging. The internal structure of the battery cell includes multiple battery cell layers. When a traditional planar imaging, such as using a CCD (charge coupled device) camera to obtain a battery cell image, cannot obtain a surface layer image of the battery cell, and cannot be used to determine the number of battery cell layers, because the traditional planar imaging does not have a penetrating property. The ray, such as an X-ray, can penetrate the battery cell, and thus the obtained image can form a pixel point with a corresponding gray value according to the number of battery cell layers penetrated by the ray. For example, if the number of battery cell layers penetrated by the ray is small, the gray value of the formed pixel point is low; otherwise, the gray value of the formed pixel point is high. In this way, the gray image of the battery cell, that is, the battery cell image representing the structure of the battery cell, can be obtained through the ray imaging, so that the gray value of the pixel in the battery cell image can be used to detect the battery cell.

[0061] When the battery cell needs to be detected, the battery cell can be subjected to ray imaging to obtain a gray image of the battery cell, which is the battery cell image. For example, the 16-bit gray image of the battery cell obtained by using an X-Ray device is used as the battery cell image, as shown in FIG. 2.

[0062] After obtaining the battery cell image, the battery cell image can be preprocessed, such as ROI (Region of Interest) extraction, to extract the battery cell region from the battery cell image, as shown in FIG. 3.

[0063] As a possible implementation, the extraction of the battery cell region can be edge extraction of the battery cell image to obtain an edge image of the battery cell image, and the battery cell region can be extracted from the battery cell image according to the edge image.

[0064] The edge extraction refers to a processing of an image contour in digital image processing. The place with the maximum gray value change rate of the image is defined as the edge, that is, the inflection point, which refers to a point where the function changes concave-convex, and the second derivative is zero. The edge extraction of the battery cell image can use an edge detection operator to perform edge detection on the battery cell image to extract the contours of all objects in the battery cell image to form an edge image. The edge detection operator can be any one of a gradient edge operator, a Roberts edge operator, a Laplacian edge operator, and a Sobel edge operator.

[0065] After the edge image is obtained, the edge image can be subjected to cell boundary detection to extract the cell region from the cell image according to the detected cell boundary. For example, an edge line corresponding to the contour of the cell can be identified from the cell image as the cell boundary to extract the cell region of the cell from the cell image using the identified cell boundary. Alternatively, the boundary of the cell can be identified from the edge image according to a preset gray value. The preset gray value can be the gray value of the boundary pixel of the cell. For example, using the preset gray value, each edge line in the edge image is traversed to find an edge line having the same gray value as the preset gray value or a difference between the gray value and the preset gray value less than a preset value from the edge image, and the cell region is extracted from the cell image using all the found edge lines.

[0066] In addition, the cell region can also be extracted from the cell image by other conventional ROI extraction methods, such as based on opencv or a pre-trained image segmentation model.

[0067] Since the cell is formed by winding the cathode pole piece, the anode pole piece and the separator, when the starting position of the winding of the cathode pole piece of the cell and the starting position of the winding of the anode pole piece, i.e., the cathode feeding position and the anode feeding position do not overlap, the number of cell layers on the region between the two is less than the number of cell layers on both sides of the region, as shown in FIG. 4a. If the cathode feeding position and the anode feeding position overlap, the number of cell layers on the region between the two is greater than the number of cell layers on both sides of the region, as shown in FIG. 4b.

[0068] If the cathode feeding position and the anode feeding position do not overlap, since the number of cell layers on the region between the cathode feeding position and the anode feeding position is less than the number of cell layers on both sides of the region, after radiographic imaging, the gray value of the region between the cathode feeding position and the anode feeding position will be greater than the gray values of the regions on both sides, as shown in FIG. 5a. Similarly, if the cathode feeding position and the anode feeding position overlap, after radiographic imaging, the gray value of the region between the cathode feeding position and the anode feeding position will be less than the gray values of the regions on both sides, as shown in FIG. 5b. Based on this, after the cell region is extracted from the cell image, the gray value distribution of the pixels in the cell region can be detected to determine whether the cathode feeding position and the anode feeding position of the cell overlap based on the gray value distribution.

[0069] For example, after the cell region is extracted, the adjacent pixels can be clustered according to the gray values of the pixels in the cell region, such as by using a KNN algorithm to obtain a plurality of sub-regions, and then the gray value distribution between the sub-regions is detected. If it is detected that the average gray value of a certain sub-region is greater than the average gray values of the sub-regions on both sides, it is determined that the cathode and anode loading positions of the cell do not overlap, and the cell is determined to be a normal cell. If it is detected that the average gray value of a certain sub-region is less than the average gray values of the sub-regions on both sides, it is determined that the cathode and anode loading positions of the cell overlap, and the cell is determined to be an abnormal cell. Alternatively, the area of a sub-region with an average gray value less than the average gray values of the sub-regions on both sides can be detected. If the area is greater than 0, it indicates that the cathode and anode loading positions overlap, and the cell is determined to be an abnormal cell. Otherwise, the cell is determined to be a normal cell.

[0070] In addition, the loading position detection result can also include the overlapping distance of the loading positions. For example, the distance between the two side boundaries of a sub-region in the cell region with an average gray value less than the average gray values of the sub-regions on both sides can be taken as the overlapping distance of the cathode and anode loading positions of the cell, so as to obtain the specific abnormal situation of the cell, facilitating subsequent classification and interception of abnormal cells. The boundary of the sub-region refers to the boundary between the sub-region and its adjacent sub-region.

[0071] The cell region is extracted from the cell image obtained by using the ray imaging, and the loading position detection result of the cell is determined according to the gray value distribution of the pixels in the cell region. When the cathode loading position overlaps with the anode loading position, the number of layers of the cell in the overlapping part increases, resulting in a large number of layers that need to be penetrated during the ray imaging. At this time, the pixel gray value of the overlapping part in the cell region is low. If the cathode loading position does not overlap with the anode loading position, the number of layers of the cell between the cathode loading position and the anode loading position is small, resulting in a small number of layers that need to be penetrated during the ray imaging. At this time, the pixel gray value between the cathode loading position and the anode loading position is high. Therefore, by detecting the gray value distribution of the pixels in the cell region, it can be determined whether the cathode loading position and the anode loading position of the finished cell overlap, thereby improving the accuracy and reliability of the cell detection result.

[0072] In order to reduce the data volume of image processing and reduce the noise interference during cell region recognition, in some embodiments, the cell region is extracted from the cell image obtained by using the ray imaging, including: performing a binaryzation processing on the cell image according to the gray values of the pixels in the cell image to obtain a binaryzation cell image of the cell image; and obtaining the cell region in the cell image according to the connected region corresponding to the cell in the binaryzation cell image.

[0073] As a possible implementation, when the battery cell region needs to be extracted from the battery cell image, the gray value of each pixel in the battery cell image can be binarized using a binarization algorithm to determine the corresponding threshold, and the battery cell image is binarized using the threshold, such as adaptive binarization, to obtain a binarized battery cell image of the battery cell image. At this time, the binarized battery cell image includes the battery cell region as the foreground and other regions as the background. The binarization algorithm can be any binarization algorithm such as the bimodal method, the P parameter method, the iterative method, and the OTSU method.

[0074] Taking the OTSU binarization algorithm as an example, when the threshold is t, P(t) is the probability of the pixel in the battery cell image being divided into the foreground pixel, the average gray value of the pixel allocated to the foreground is μ1(t), [1-P(t)] is the probability of the pixel in the battery cell image being divided into the background, and the average gray value is μ0(t). The cumulative average of the gray level t is μ total , and the gray value of the entire battery cell image is μ g . Therefore, P(t)*μ1(t)+[1-P(t)]*μ0(t)=μ g (1)

[0075] According to the variance concept, the expression can be expressed as: σ 2 (t)=P(t)·(μ1(t)-μ g ) 2 +[1-P(t)]·(μ0(t)-μ g ) 2 (2)

[0076] Substituting formula (1) into formula (2), σ 2 (t)=P(t)·[1-P(t)]·(μ1(t)-μ0(t)) 2 (3)

[0077] Finally, it can be written as the following formula, where

[0078] After solving the optimal threshold value using the OTSU algorithm, the battery cell image can be binarized according to the optimal threshold value:

[0079] Where (i,j) represents the coordinates of the pixel in the battery cell image.

[0080] To further improve the clarity of the binarized battery cell image, the image can be denoised again after adaptive binarization, such as through the open operation to denoise the image after adaptive binarization to obtain a binarized battery cell image of the battery cell image.

[0081] After obtaining the binarized battery image, a connected region corresponding to the battery can be extracted from the binarized battery image, and a region corresponding to the connected region in the battery image can be extracted as the battery region.

[0082] For example, the binarized battery image can be subjected to connected region analysis, that is, each pixel in the binarized battery image is assigned a new label according to the labels of its neighboring pixels. Specifically, new_label(i,j) = min(neighbor_labels(i,j))

[0083] where new_label(i,j) is the new label of pixel (i,j), and neighbor_labels(i,j) is a set of labels of neighboring pixels of pixel (i,j). The labels include foreground or background, and the initial label of a pixel is determined by the binarization described above.

[0084] Then, the area and the circumscribed rectangle of each connected region are counted. Specifically, area(label) = Σ i,j [label(i,j)==label]

[0085] where area(label) represents the area of the region with label label.

[0086] After obtaining the connected regions, the largest region in the connected regions is determined as the connected region corresponding to the battery, and a region corresponding to the connected region in the battery image is extracted as the battery region.

[0087] By binarizing the battery image, a binarized battery image of the battery image is obtained, and based on the connected region corresponding to the battery in the binarized battery image, the battery region in the battery image is obtained. Thus, the noise in the battery image can be effectively removed by binarization, and the battery region can be accurately extracted from the battery image by using the connected region in the binarized image, thereby improving the accuracy of battery region identification.

[0088] To further improve the accuracy of battery region identification, in some embodiments, the battery image is binarized according to the gray values of the pixels in the battery image to obtain a binarized battery image of the battery image, including: binarizing the battery image according to the gray values of the pixels in the battery image to obtain an initial binarized image of the battery image; and sequentially performing image erosion and image dilation on the initial binarized image to obtain the binarized battery image of the battery image.

[0089] In some embodiments, after the battery cell image is binarized according to the gray value of each pixel in the battery cell image, an initial binarized image of the battery cell image can be obtained. After the initial binarized image is obtained, the initial binarized image can be eroded and then dilated, so as to effectively remove small objects, isolated regions and thin edges, while keeping the shape of the larger connected regions in the image.

[0090] For example, after the initial binarized battery cell image is obtained through the binarization process, a small kernel called a structural element can be used to slide over the initial binarized battery cell image and perform convolution. For each pixel in the initial binarized battery cell image, the pixels in the region covered by the structural element are updated according to certain rules, such as taking the minimum pixel value as the new value of the center pixel. Assuming that the input image, i.e., the initial binarized battery cell image, is A, the structural element is B, and the erosion operation is denoted by The calculation formula is as follows:

[0091] where (i, j) represents the pixel coordinates in the image A, (k, l) represents the element coordinates in the structural element B, and the above formula indicates that the minimum value of all covered pixels is taken as the new pixel value within the range of the structural element.

[0092] Then, the initial binarized image after the image erosion operation is dilated. The image dilation can expand or dilate the objects in the image by sliding the structural element. The dilation operation is denoted by The calculation formula is as follows:

[0093] The calculation formula of the opening operation is as follows:

[0094] In this way, after the image erosion and the image dilation, the binarized battery cell image of the battery cell image Opening(A) can be obtained, so as to effectively denoise the binarized image of the battery cell image, improve the clarity and readability of the obtained binarized image, and further improve the accuracy of the battery cell region extracted from the binarized image.

[0095] After the binarized battery cell image is obtained, the battery cell region corresponding to the connected region corresponding to the battery cell in the binarized battery cell image can be extracted from the battery cell image.

[0096] To improve the clarity and readability of the battery cell region extracted from the battery cell image, and to improve the accuracy of detecting the battery cell region, in some embodiments, the battery cell region in the battery cell image is obtained according to a connected region corresponding to the battery cell in the binary battery cell image, including: obtaining a target region in the battery cell image according to the connected region; and performing image enhancement processing on the target region to obtain the battery cell region; wherein the image enhancement processing includes performing filtering processing on each column of pixels in the target region located in the vertical direction of the inlet position.

[0097] In some embodiments, after obtaining the connected region corresponding to the battery cell in the binary battery cell image, the region corresponding to the connected region in the battery cell image can be extracted as the target region according to the connected region.

[0098] After obtaining the target region, since the features of the inlet position of the battery cell in the battery cell image obtained by the radiographic imaging are all reflected in the vertical direction of the cathode and anode inlet positions, at least the filtering processing can be performed on each column of pixels in the target region located in the vertical direction of the cathode and anode inlet positions of the battery cell, to highlight or enhance the features in the vertical direction, so that the features of the cathode and anode inlet positions in the obtained battery cell region are more prominent, and the cathode inlet position and the anode inlet position can be more clearly observed, facilitating subsequent battery cell detection.

[0099] In addition to performing the filtering processing on each column of pixels in the target region located in the vertical direction of the inlet position, at least one image enhancement processing mode such as global contrast stretching, Laplacian filtering, and Gaussian filtering can be used to perform image enhancement on the target region.

[0100] As a possible implementation, the image enhancement processing on the target region can be adjusting the gray level distribution of the target region using global contrast stretching; performing filtering processing on each column of pixels in the target region located in the vertical direction of the inlet position to highlight or enhance the pixel features of each column of pixels located in the vertical direction of the inlet position; performing Laplacian filtering on the target region to highlight the edges and details in the target region to enhance the image features; and performing Gaussian filtering on the target region after the Laplacian filtering to reduce some high-frequency noise introduced by the Laplacian filtering, smooth the image, improve the overall contrast of the image, and reduce the phenomenon of discontinuous image edges that can occur after the Laplacian filtering, thereby obtaining the battery cell region.

[0101] For example, after obtaining the target region, the pixel value range of the target region can be adjusted to enhance the image contrast of the target region. For example, the pixel value of the target region is adjusted from the original range [I min ,I maxextends to new range [0, 65535], where I(x, y) is the pixel value of the target region, I'(x, y) is the pixel value after contrast stretching, I min is the minimum pixel value of the target region, I max is the maximum pixel value of the target region.

[0102] After completing the global contrast stretching of the target region, the target region after the global contrast stretching can be subjected to vertical filtering, i.e. filtering processing is performed on each column of pixels in the vertical direction of the input position. A one-dimensional vertical filter is defined, where the one-dimensional filter (kernel) is h, and the length of the filter is L, k represents the position offset of the filter in the vertical direction of the input image, i.e. the target region, x[x, y+k] is the pixel value of the target region, and y[x, y'] is the pixel value of the target region after vertical filtering.

[0103] After completing the vertical filtering of the target region, the target region after the vertical filtering can be subjected to Laplacian filtering. The target region is subjected to twice differentiation, so as to detect the edges and details in the target region.

[0104] I yy (x, y) = I(x, y+1) - 2I(x, y) + I(x, y-1) then the Laplacian operator calculation formula is as follows:

[0105] After completing the Laplacian filtering of the target region by using the Laplacian operator, the target region after the Laplacian filtering can be subjected to Gaussian filtering. The pixel value of the target region is subjected to weighted average, the target region is subjected to smoothing processing and noise removal, and the image details and edges are retained at the same time.

[0106] wherein the Gaussian function is as follows:

[0107] After the target region completes the above image enhancement processing, the target region after the image enhancement processing can be subjected to region of interest extraction again, so as to obtain the battery cell region. The target region after the above image enhancement processing is determined as the battery cell region. The finally extracted battery cell region can be as shown in FIG. 5a or FIG. 5b. In this way, the features of the anode and cathode input positions in the obtained battery cell region are more prominent, which facilitates subsequent battery cell detection and improves the accuracy of battery cell detection.

[0108] After the extraction of the battery cell region, the distribution of the gray values of the pixels in the battery cell region can be used to determine the loading of the battery cell as the detection result. Considering that the loading position of the battery cell in the image obtained by the radiographic imaging is reflected in the vertical direction of the loading position of the cathode and the anode, in some embodiments, the loading position detection result of the battery cell is obtained according to the distribution of the gray values of the pixels in the battery cell region, including: obtaining the loading position detection result of the battery cell according to the gray values of the pixels in each column in the vertical direction of the loading position in the battery cell region.

[0109] The gray value of any column of pixels can be the average gray value of the pixel points in the column of pixels or the mode value in the gray values of the pixel points in the column of pixels. For example, for any column of pixels, the gray values of the pixel points in the column of pixels are added to obtain the average value, and the average value of the pixels in the column of pixels is obtained, that is, the gray value of the column of pixels.

[0110] wherein I represents the battery cell region, H represents the height of the battery cell region, I[i,j] represents the gray value of the pixel point in the i-th row and the j-th column of the battery cell region, and P[j] represents the average value of the pixels in the j-th column, that is, the gray value of the j-th column of pixels.

[0111] When the loading position of the cathode and the loading position of the anode do not overlap, the number of layers of the radiographic penetration at the loading positions of the cathode and the anode decreases by two, as shown in FIG. 4a. At this time, the higher the brightness after imaging, the higher the gray value of the pixels between the loading positions of the cathode and the anode. Conversely, the number of layers of the radiographic penetration between the loading positions of the cathode and the anode increases by two, as shown in FIG. 4b. At this time, the lower the brightness after imaging, the lower the gray value of the pixels between the loading positions of the cathode and the anode. Therefore, after obtaining the gray values of the pixels in each column in the vertical direction of the loading position, each column of pixels can be traversed from left to right or from right to left to obtain the gray value change trend of each column of pixels. If there are some column of pixels with a gray value change trend from small to large and then to small in each column of pixels, it can be determined that the loading position of the cathode and the loading position of the anode do not overlap. If there are some column of pixels with a gray value change trend from large to small and then to large in each column of pixels, it can be determined that the loading position of the cathode and the loading position of the anode overlap.

[0112] Alternatively, the minimum gray value of each pixel in the battery cell region of an abnormal battery cell in which the cathode inlet position and the anode inlet position overlap can be obtained as a preset gray value. The abnormal battery cell has the same specifications and model as the battery cell to be detected. It can be understood that the abnormal battery cell region is obtained in the same manner as the battery cell region of the battery cell to be detected, and the above-described manner can be used for obtaining. Since the number of layers of the ray that needs to be penetrated between the cathode inlet position and the anode inlet position is the largest when the cathode inlet position and the anode inlet position overlap, the gray value of the pixel located between the cathode inlet position and the anode inlet position is the smallest, and thus the preset gray value can represent the gray value when the cathode inlet position and the anode inlet position overlap.

[0113] After obtaining the gray values of the pixels in each column, a plurality of boundaries can be determined from the pixels in each column according to the difference between the gray values of adjacent pixels in each column. For example, if the difference between the gray values of two adjacent pixels in each column reaches a preset value, the pixel with the smaller or larger gray value in each column can be extracted as a boundary. The preset value can be set according to actual conditions. After a plurality of boundaries are extracted, the gray values of the pixels in each column between any two boundaries can be detected to determine whether the gray values match the preset gray value, for example, whether the difference between the gray values and the preset gray value is within a preset range. If the gray values of the pixels in each column between the two boundaries match the preset gray value, for example, the difference between the gray values of the pixels in each column between the two boundaries and the preset gray value is less than a preset threshold, it can be determined that the battery cell inlet position detection result is that the inlet positions overlap. Otherwise, it can be determined that the battery cell inlet position detection result is that the inlet positions do not overlap. The preset threshold can be set according to actual conditions. For example, a plurality of preset gray values corresponding to a plurality of abnormal battery cells having the same specifications and model as the battery cell to be detected can be obtained, and then the difference between the maximum value and the minimum value of each preset gray value can be used as the preset threshold.

[0114] The gray values of the pixels in each column in the vertical direction of the battery cell region are used to determine the battery cell inlet position detection result, so that the inlet position characteristics of the battery cell can be effectively used for detection when the battery cell is detected, and the accuracy of the battery cell inlet position detection result is improved.

[0115] In some embodiments, the battery cell inlet position detection result is obtained according to the gray values of the pixels in each column in the vertical direction of the battery cell region, which includes:

[0116] The gray values of the pixels in each column are compared with a preset gray value to obtain the battery cell inlet position detection result. The preset gray value is the minimum gray value detected from the battery cell region of an abnormal battery cell in which the inlet positions overlap.

[0117] In some embodiments, the minimum gray value of each pixel in the abnormal battery cell can be obtained as the preset gray value, where the abnormal battery cell has the same specification and model as the battery cell to be detected.

[0118] After obtaining the gray values of each column of pixels, the gray values of each column of pixels can be compared with the preset gray value. If there is a column of pixels with a gray value less than or equal to the preset gray value, or there is a column of pixels with a gray value difference from the preset gray value less than the preset threshold, it can be determined that the battery cell has the overlapping of the material inlet position. If the gray values of each column of pixels are all greater than the preset gray value, or the gray values of each column of pixels are all greater than the preset gray value, and the gray value difference of each column of pixels from the preset gray value is greater than the preset threshold, it can be determined that the battery cell does not have the overlapping of the material inlet position. In this way, the gray values of each column of pixels can be compared with the preset gray value to quickly determine whether the battery cell has the overlapping of the material inlet position, thereby improving the detection efficiency of the battery cell.

[0119] To make the detection result of the material inlet position of the battery cell more accurate, in addition to comparing the gray values of each column of pixels with the preset gray value to obtain the detection result of the material inlet position of the battery cell, in some embodiments, as shown in FIG. 6, the detection result of the material inlet position of the battery cell is obtained according to the gray values of each column of pixels in the vertical direction of the material inlet position in the battery cell region, including:

[0120] S201, obtaining the gray values of each column of pixels in the vertical direction of the material inlet position in the battery cell region;

[0121] S202, obtaining the gray value change data of the battery cell region according to the gray values of adjacent columns of pixels;

[0122] S203, obtaining the detection result of the material inlet position of the battery cell according to the gray value change data.

[0123] In some embodiments, after obtaining the battery cell region, the gray values of each column of pixels can be extracted in the vertical direction of the material inlet position from left to right or from right to left, and then the gray values of adjacent columns of pixels can be arranged in sequence, so as to obtain the gray value change data of the battery cell region. Alternatively, the gray values of adjacent columns of pixels can be converted into a gray value change curve representing the corresponding relationship between the gray value and the position of the battery cell, taking the left side of the battery cell region as the starting position. The gray value change curve is the gray value change data of the battery cell region.

[0124] As for a normal battery cell with no overlap between the cathode loading position and the anode loading position, based on the principle of ray imaging, such as the principle of X-ray imaging, the number of layers penetrated by the ray between the cathode loading position and the anode loading position is reduced, so the gray level change data of the normal battery cell will suddenly rise to form a peak, and then suddenly drop, as shown in FIG. 7. As for an abnormal battery cell with overlap between the cathode loading position and the anode loading position, based on the principle of ray imaging, the number of layers penetrated by the ray between the cathode loading position and the anode loading position is increased, so the gray level change data of the abnormal battery cell will suddenly drop to form a trough, and then suddenly rise, as shown in FIG. 8.

[0125] Therefore, after obtaining the gray level change data of the battery cell region, the change trend of the gray level change data of the battery cell region can be detected. For example, the gray level change data of the battery cell region is matched with the gray level change data of a normal battery cell in terms of similarity. If the change trend of the gray level change data of the battery cell region is consistent with that of the gray level change data of the normal battery cell, such as the similarity between the two is greater than a preset similarity, such as 80%, it can be determined that the loading position detection result of the battery cell corresponding to the battery cell region is that the loading positions are not overlapped; otherwise, it can be determined that the loading position detection result of the battery cell corresponding to the battery cell region is that the loading positions are overlapped. Alternatively, the gray level change data of the battery cell region is matched with the gray level change data of an abnormal battery cell in terms of similarity. If the change trend of the gray level change data of the battery cell region is consistent with that of the gray level change data of the abnormal battery cell, such as the similarity between the two is greater than a preset similarity, it can be determined that the loading position detection result of the battery cell corresponding to the battery cell region is that the loading positions are overlapped; otherwise, it can be determined that the loading position detection result of the battery cell corresponding to the battery cell region is that the loading positions are not overlapped.

[0126] Alternatively, the preset model can also be used to detect the gray level change data to obtain the loading position detection result of the battery cell. The preset model can be an SVM classifier or other lightweight AI detection model.

[0127] For training of the preset model, a large number of battery cell regions can be extracted from normal battery cells and abnormal battery cells as battery cell region samples. Then, the gray level change data of each battery cell region sample is determined in the above manner, and then input into the preset model for training to obtain the pre-trained preset model.

[0128] As a possible implementation, the gray level change data of each battery cell region sample can be sequentially input into the preset model. After obtaining the predicted loading position detection result of whether the preset model overlaps each time, the predicted loading position is matched with the preset loading position detection result corresponding to the gray level change data input this time.

[0129] In the case of mismatch, the network parameters of the preset model are adjusted by using the gradient descent method through error back propagation, and the next training is performed until the prediction of the charging level detection result obtained after inputting the gray level change data of the cell region sample each time matches the preset charging level detection result corresponding to the input gray level change data, indicating that the training of the preset model is completed, and the pre-trained preset model is obtained.

[0130] In this way, when judging the charging level detection result of the cell, the overall distribution of each column of pixels in the vertical direction of the charging level is considered, and the accuracy of the charging level detection result is improved.

[0131] In some embodiments, in addition to judging whether the cathode charging level and the anode charging level of the cell overlap, the charging level detection result can also include the overlap distance between the anode charging level and the cathode charging level of the cell. Wherein, the charging level detection result of the cell is obtained according to the gray level change data, including: determining the anode charging level and the cathode charging level of the cell according to the gray level difference of adjacent gray levels in the gray level change data; obtaining the overlap distance between the anode charging level and the cathode charging level of the cell according to the anode charging level and the cathode charging level.

[0132] In some embodiments, after obtaining the gray level change data of the cell region, the gray level difference of adjacent gray levels in the gray level change data can be obtained first. The gray level difference is the absolute value of the difference between adjacent gray levels. Then, one of the two columns of pixels corresponding to the highest gray level difference in each gray level difference and one of the two columns of pixels corresponding to the second highest gray level difference in each gray level difference are determined as the anode charging level and the cathode charging level of the cell, respectively.

[0133] After determining the anode charging level and the cathode charging level of the cell, if the gray level change data of the cell region matches the gray level change data of the normal cell, or does not match the gray level change data of the abnormal cell, the overlap distance between the anode charging level and the cathode charging level of the cell is determined as 0; if the gray level change data of the cell region matches the gray level change data of the abnormal cell, or does not match the gray level change data of the normal cell, the distance between the anode charging level and the cathode charging level is determined as the overlap distance between the anode charging level and the cathode charging level of the cell.

[0134] Alternatively, after determining that the gray scale change data of the battery cell region matches the gray scale change data of the abnormal battery cell or does not match the gray scale change data of the normal battery cell, one of the two columns of pixels corresponding to the highest gray scale difference in each gray scale difference and one of the two columns of pixels corresponding to the second highest gray scale difference in each gray scale difference are determined as the anode charging position and the cathode charging position of the battery cell, respectively, so that the distance between the anode charging position and the cathode charging position is determined as the overlap distance between the anode charging position and the cathode charging position of the battery cell.

[0135] Alternatively, after determining the anode charging position and the cathode charging position of the battery cell, it can be detected whether the gray scale values of each column of pixels between the anode charging position and the cathode charging position in the gray scale change data are less than the gray scale values of each column of pixels outside the anode charging position and the cathode charging position; if so, the distance between the anode charging position and the cathode charging position is determined as the overlap distance between the anode charging position and the cathode charging position of the battery cell; otherwise, the overlap distance between the anode charging position and the cathode charging position of the battery cell is determined as 0.

[0136] The anode charging position and the cathode charging position of the battery cell are determined through the gray scale difference between adjacent gray scale values in the gray scale change data, and the overlap distance between the anode charging position and the cathode charging position of the battery cell is obtained based on the anode charging position and the cathode charging position, so that the obtained charging position detection result is more accurate and facilitates subsequent adjustment of the battery cell.

[0137] In addition, in some embodiments, after obtaining the gray scale change data of the battery cell region, the gray scale change data can also be detected using a pre-trained preset model to obtain the overlap distance between the anode charging position and the cathode charging position of the battery cell; wherein the preset model is trained by the gray scale change data corresponding to each battery cell region sample, and each battery cell region sample includes a battery cell region of a normal battery cell and a battery cell region of an abnormal battery cell.

[0138] As a possible implementation, the gray scale change data corresponding to each battery cell region sample can be sequentially input into the preset model, and each time the input is performed, the predicted overlap distance between the anode charging position and the cathode charging position output by the preset model is obtained, and then the predicted overlap distance is matched with the preset overlap distance corresponding to the gray scale change data input this time.

[0139] In the case where the two do not match, the gradient descent method is used to adjust the network parameters of the preset model through error back propagation, and the next training is performed until the predicted overlap distance obtained after each input of the gray scale change data matches the preset overlap distance corresponding to the training sample input this time, which indicates that the training of the preset model is completed, and the pre-trained preset model is obtained.

[0140] After the pre-trained preset model is acquired, the gray level change data of the battery cell region can be input into the pre-trained preset model, so as to identify the overlap distance between the anode filling position and the cathode filling position of the battery cell through the pre-trained preset model.

[0141] For the purpose, technical solutions and advantages of the present application to be more clear, the technical solutions in the present application will be described clearly and completely. In some embodiments, as shown in FIG. 9, the battery cell detection method comprises:

[0142] S301, acquiring a battery cell image generated by radiographic imaging of a battery cell;

[0143] S302, performing binaryzation processing on the battery cell image according to the gray level values of each pixel in the battery cell image, to obtain an initial binaryzation image of the battery cell image;

[0144] S303, performing erosion operation and inflation operation on the initial binaryzation image in sequence, to obtain a binaryzation battery cell image of the battery cell image;

[0145] S304, performing connected region analysis on the binaryzation battery cell image, extracting a region with the largest area as a region of interest, to determine a target region in the battery cell image according to the region of interest;

[0146] S305, adjusting the gray level distribution of the target region using global contrast stretching, and performing longitudinal filtering, Laplace filtering and Gaussian filtering on the target region after global contrast stretching, to obtain a battery cell region;

[0147] S306, obtaining gray level change data of the battery cell region according to the gray level values of each column of pixels adjacent to each other in the vertical direction of the filling position in the battery cell region;

[0148] S307, inputting the gray level change data of the battery cell region into a pre-trained SVM classifier for detection, to obtain the overlap distance between the anode filling position and the cathode filling position of the battery cell.

[0149] FIG. 10 shows a schematic structural block diagram of an electric cell detection device according to an embodiment of the present application. It should be understood that the device corresponds to the method embodiments performed in FIG. 1, FIG. 6 and FIG. 9, and can perform the steps involved in the aforementioned methods. The specific functions of the device can be referred to the description above, and the detailed description is appropriately omitted here to avoid repetition. The device comprises at least one software function module stored in the form of software or firmware in the memory or solidified in the operating system (OS) of the device. Specifically, the device comprises: a region extraction module 401 configured to extract an electric cell region from an electric cell image obtained by radiographic imaging; and an electric cell detection module 402 configured to obtain an electric cell feeding position detection result according to the gray value distribution of the pixels in the electric cell region.

[0150] In the technical solution of the embodiments of the present application, the electric cell region is extracted from the electric cell image obtained by radiographic imaging, so as to determine the electric cell feeding position detection result according to the gray value distribution of the pixels in the electric cell region. When the cathode feeding position and the anode feeding position overlap, the number of layers of the electric cell in the overlapping part increases, resulting in a large number of layers to be penetrated during radiographic imaging, and the pixel gray value of the overlapping part in the electric cell region is low at this time. If the cathode feeding position and the anode feeding position do not overlap, the number of layers of the electric cell between the cathode feeding position and the anode feeding position is small, resulting in a small number of layers to be penetrated during radiographic imaging, and the pixel gray value between the cathode feeding position and the anode feeding position is high at this time. Therefore, by detecting the gray value distribution of the pixels in the electric cell region, it can be determined whether the cathode feeding position and the anode feeding position of the finished electric cell overlap, thereby improving the accuracy and reliability of the electric cell detection result.

[0151] According to some embodiments of the present application, the region extraction module 401 is specifically configured to: perform binaryzation processing on the electric cell image according to the gray values of the pixels in the electric cell image, to obtain a binaryzation electric cell image of the electric cell image; and obtain the electric cell region in the electric cell image according to the connected region corresponding to the electric cell in the binaryzation electric cell image.

[0152] According to some embodiments of the present application, the region extraction module 401 is specifically configured to: perform binaryzation processing on the electric cell image according to the gray values of the pixels in the electric cell image, to obtain an initial binaryzation image of the electric cell image; and sequentially perform image erosion and image dilation on the initial binaryzation image, to obtain a binaryzation electric cell image of the electric cell image.

[0153] According to some embodiments of the present application, the region extraction module 401 is specifically configured to: obtain a target region in the battery cell image according to the connected region; and perform image enhancement processing on the target region to obtain the battery cell region; wherein the image enhancement processing comprises performing filtering processing on each column of pixels in the target region in the vertical direction of the charging position.

[0154] According to some embodiments of the present application, the battery cell detection module 402 is specifically configured to:

[0155] obtain a charging position detection result of the battery cell according to the gray scale values of each column of pixels in the battery cell region in the vertical direction of the charging position.

[0156] According to some embodiments of the present application, the battery cell detection module 402 is specifically configured to:

[0157] compare the gray scale values of each column of pixels with a preset gray scale value to obtain a charging position detection result of the battery cell; wherein the preset gray scale value is a minimum gray scale value detected from the battery cell region of the abnormal battery cell overlapping the charging position.

[0158] According to some embodiments of the present application, the battery cell detection module 402 is specifically configured to:

[0159] obtain the gray scale values of each column of pixels in the battery cell region in the vertical direction of the charging position; obtain gray scale change data of the battery cell region according to the gray scale values of adjacent columns of pixels; and obtain a charging position detection result of the battery cell according to the gray scale change data.

[0160] According to some embodiments of the present application, the battery cell detection module 402 is specifically configured to: determine the anode charging position and the cathode charging position of the battery cell according to the gray scale difference between adjacent gray scale values in the gray scale change data; and obtain an overlapping distance between the anode charging position and the cathode charging position of the battery cell according to the anode charging position and the cathode charging position.

[0161] According to some embodiments of the present application, the battery cell detection module 402 is specifically configured to: use a pre-trained preset model to detect the gray scale change data to obtain an overlapping distance between the anode charging position and the cathode charging position of the battery cell; wherein the preset model is trained by gray scale change data corresponding to each battery cell region sample, and each battery cell region sample comprises a battery cell region of a normal battery cell and a battery cell region of an abnormal battery cell.

[0162] According to some embodiments of the present application, as shown in FIG. 11, the electronic device 5 comprises a processor 501 and a memory 502, the processor 501 and the memory 502 are interconnected and communicate with each other through a communication bus 503 and / or other forms of connection mechanism (not shown), the memory 502 stores a computer program executable by the processor 501, when the computing device is running, the processor 501 executes the computer program to execute the method performed by the external terminal in any optional implementation manner, for example: extracting the battery region from the battery image obtained by the radiographic imaging; obtaining the charging position detection result of the battery according to the gray value distribution of the pixels in the battery region.

[0163] The present application provides a computer readable storage medium, the computer readable storage medium stores a computer program, the computer program is executed by a processor to execute the method in any optional implementation manner.

[0164] The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0165] The present application provides a computer program product, which, when running on a computer, causes the computer to execute the method in any optional implementation manner.

[0166] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than limit them. Although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some or all of the technical features can be replaced equivalently. Such modifications or replacements do not change the essence of the corresponding technical solutions, which should be covered in the scope of the claims and the specification of the present application. In particular, the technical features mentioned in each embodiment can be combined in any manner as long as there is no structural conflict. The present application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method of detecting a battery cell, the method comprising: The method comprises: extracting the battery cell region from the battery cell image obtained through the ray imaging; obtaining the feeding position detection result of the battery cell according to the gray value distribution of the pixels in the battery cell region.

2. The method of claim 1, wherein, The method comprises: extracting the battery cell region from the battery cell image obtained through the ray imaging, comprising: performing binaryzation processing on the battery cell image according to the gray value of each pixel in the battery cell image to obtain a binaryzation battery cell image of the battery cell image; 3. The method of claim 2, wherein, obtaining the battery cell region in the battery cell image according to the connected region corresponding to the battery cell in the binaryzation battery cell image. The method comprises: performing binaryzation processing on the battery cell image according to the gray value of each pixel in the battery cell image to obtain a binaryzation battery cell image of the battery cell image, comprising:

4. The method according to claim 2 or 3, characterized in that, performing binaryzation processing on the battery cell image according to the gray value of each pixel in the battery cell image to obtain an initial binaryzation image of the battery cell image; performing image erosion and image dilation on the initial binaryzation image in sequence to obtain the binaryzation battery cell image of the battery cell image. The method comprises: obtaining the battery cell region in the battery cell image according to the connected region corresponding to the battery cell in the binaryzation battery cell image, comprising:

5. The method of claim 4, wherein, obtaining a target region in the battery cell image according to the connected region; performing image enhancement processing on the target region to obtain the battery cell region; 6. The method of claim 1, wherein, wherein the image enhancement processing comprises performing filtering processing on each column of pixels in the target region located in the vertical direction of the feeding position. The method comprises: performing global contrast stretching, Laplacian filtering, Gaussian filtering and filtering processing on each column of pixels in the target region located in the vertical direction of the feeding position to obtain the battery cell region.

7. The method according to any one of claims 1-3, 5 or 6, characterized in that, The method comprises: performing edge extraction on the battery cell image to obtain an edge image of the battery cell image; 8. The method of claim 7, wherein, extracting the battery cell region from the battery cell image according to the edge image. The method comprises: obtaining the feeding position detection result of the battery cell according to the gray value of each column of pixels in the battery cell region located in the vertical direction of the feeding position.

9. The method of claim 7, wherein, The method comprises: comparing the gray value of each column of pixels with a preset gray value to obtain the feeding position detection result of the battery cell; wherein the preset gray value is the minimum gray value detected from the battery cell region of the abnormal battery cell overlapping the feeding position. The method comprises: obtaining the feeding position detection result of the battery cell according to the gray value of each column of pixels in the battery cell region located in the vertical direction of the feeding position, comprising: obtaining the gray value of each column of pixels in the battery cell region located in the vertical direction of the feeding position; obtaining gray value change data of the battery cell region according to the gray value of adjacent columns of pixels; obtaining the feeding position detection result of the battery cell according to the gray value change data.

10. The method of claim 9, wherein, According to the gray value change data, an electrode material filling position detection result of the battery cell is obtained, including: According to the gray value difference of adjacent gray values in the gray value change data, the anode electrode material filling position and the cathode electrode material filling position of the battery cell are determined; According to the anode electrode material filling position and the cathode electrode material filling position, an overlap distance between the anode electrode material filling position and the cathode electrode material filling position of the battery cell is obtained.

11. The method of claim 10, wherein, According to the gray value change data, an electrode material filling position detection result of the battery cell is obtained, including: The gray value change data is detected using a pre-trained preset model to obtain an overlap distance between the anode electrode material filling position and the cathode electrode material filling position of the battery cell; The preset model is trained by the gray value change data corresponding to each battery cell region sample, and each battery cell region sample includes a battery cell region of a normal battery cell and a battery cell region of an abnormal battery cell.

12. The method of any one of claims 1-3, 5, or 6, wherein, According to the gray value distribution of the pixels in the battery cell region, an electrode material filling position detection result of the battery cell is obtained, including: According to the gray values of the pixels in the battery cell region, adjacent pixels are clustered to obtain a plurality of sub-regions; According to the average gray values of the sub-regions, the electrode material filling position detection result of the battery cell is obtained.

13. The method of claim 12, wherein, According to the average gray values of the sub-regions, the electrode material filling position detection result of the battery cell is obtained, including: The average gray values of the sub-regions are detected, and in a case where the average gray value of a sub-region is less than the average gray values of the sub-regions on both sides of the sub-region, it is determined that the electrode material filling position detection result of the battery cell is that the anode electrode material filling position and the cathode electrode material filling position of the battery cell overlap.

14. The method of claim 13, wherein, The method further includes: According to the distance between the two side boundaries of the sub-region whose average gray value is less than the average gray values of the sub-regions on both sides in the battery cell region, an overlap distance of the anode electrode material filling position and the cathode electrode material filling position of the battery cell is obtained; The boundary of the sub-region is the boundary between the sub-region and the adjacent sub-region thereof.

15. The method of claim 12, wherein, According to the average gray values of the sub-regions, the electrode material filling position detection result of the battery cell is obtained, including: The average gray values of the sub-regions are detected, and in a case where the average gray value of any of the sub-regions is greater than the average gray values of the sub-regions on both sides of the sub-region, it is determined that the electrode material filling position detection result of the battery cell is that the anode electrode material filling position and the cathode electrode material filling position of the battery cell do not overlap.

16. The method of claim 12, wherein, According to the average gray values of the sub-regions, the electrode material filling position detection result of the battery cell is obtained, including: The average gray values of the sub-regions are obtained; The area of the sub-region whose average gray value is less than the average gray values of the sub-regions on both sides is detected; In a case where the area of the sub-region is greater than a preset area, it is determined that the electrode material filling position detection result of the battery cell is that the anode electrode material filling position and the cathode electrode material filling position of the battery cell overlap.

17. A method of detecting a cell, the method comprising: The method includes: An image of a battery cell generated by radiographic imaging is obtained; According to the gray values of the pixels in the image of the battery cell, a binaryzation process is performed on the image of the battery cell to obtain an initial binaryzation image of the image of the battery cell; The initial binaryzation image is sequentially subjected to an erosion operation and a dilation operation to obtain a binaryzation battery cell image of the image of the battery cell; A connected region analysis is performed on the binaryzation battery cell image, and a region with the largest connected region area is extracted as a region of interest, so as to determine a target region in the image of the battery cell according to the region of interest; The target region is adjusted in gray scale distribution using global contrast stretching, and the target region after global contrast stretching is filtered longitudinally, Laplacian filtered and Gaussian filtered to obtain a battery cell region; Gray scale variation data of the battery cell region is obtained according to gray scale values of each column of pixels adjacent to each other in the vertical direction of the feeding position in the battery cell region; The feeding position detection result of the battery cell is obtained according to the gray scale variation data.

18. An electric cell detection apparatus, characterized by comprising: The device comprises: a region extraction module configured to extract a battery cell region from a battery cell image obtained through radiographic imaging; a battery cell detection module configured to obtain a feeding position detection result of a battery cell according to a gray scale value distribution of pixels in the battery cell region.

19. An electronic device comprising a processor and a memory having a computer program stored therein, characterized in that, The processor executes the computer program to implement the method of any one of claims 1 to 17.

20. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method of any one of claims 1 to 17.

Images