Battery cell detection method, electronic equipment and battery cell detection equipment

Abstract

The invention discloses a battery cell detection method, electronic equipment and battery cell detection equipment. The method comprises the following steps: acquiring a to-be-detected image of a battery cell; detecting the to-be-detected image based on a plurality of predefined anchor frames to obtain a first target area; performing multiple convolution operations and deconvolution operations on the to-be-detected image to obtain a second target area; and determining a union set of the first target area and the second target area as a detected target area set. The method comprises the following steps: detecting a to-be-detected image based on a plurality of predefined anchor frames to obtain a first target area; and performing multiple convolution operations and deconvolution operations on the to-be-detected image to obtain a second target region, and determining the union set of the first target region and the second target region as the target region set obtained by detection, so that all foreign matters in the battery cell can be accurately detected, thereby improving the comprehensiveness and accuracy of detection, and improving the detection efficiency. And the problems of cell performance reduction and potential safety hazards caused by foreign matters are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of battery cell detection, and particularly to a battery cell detection method, an electronic device, and a battery cell detection device. Background Art

[0002] With the rapid development of the new energy vehicle industry, as the core component of new energy vehicles, the safety and stability requirements of power batteries are getting higher and higher. As the basic unit of power batteries, the internal quality of battery cells directly affects the overall performance of the battery. Therefore, in the production process of battery cells, it is particularly important to perform non-destructive testing on them.

[0003] Currently, in the field of battery cell detection, X-ray detection equipment is mainly used to detect the alignment of the anode and cathode of battery cells. This method can monitor the alignment of each layer of battery cells in real time during the production and manufacturing process to ensure the structural stability of the battery cells.

[0004] However, although the existing detection methods can detect the alignment of battery cells, they do not identify foreign objects between each layer inside the battery cells. These foreign objects may accidentally fall in during the production process, potentially affecting the performance and safety of the battery cells. Summary of the Invention

[0005] Aiming at the deficiencies of the existing technology, the present application provides a battery cell detection method, an electronic device, and a battery cell detection device. By detecting a to-be-detected image based on a plurality of predefined anchor boxes, a first target region of the to-be-detected image is obtained; performing multiple convolutional operations and deconvolution operations on the to-be-detected image to classify the objects in the to-be-detected image, a second target region of the to-be-detected image is obtained, and the union of the first target region and the second target region is determined as the detected target region set, which can accurately detect all foreign objects in the battery cell, thereby improving the comprehensiveness and accuracy of the detection, and avoiding problems such as the performance degradation of the battery cell and potential safety hazards caused by foreign objects.

[0006] To solve the above problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present application provides a battery cell detection method, including: obtaining a to-be-detected image of a battery cell;

[0008] detecting the to-be-detected image based on a plurality of predefined anchor boxes to obtain a first target region of the to-be-detected image;

[0009] performing multiple convolutional operations and deconvolution operations on the to-be-detected image to classify the objects in the to-be-detected image, and obtaining a second target region of the to-be-detected image, where the objects in the to-be-detected image include battery cells and foreign objects;

[0010] Determine the union of the first target region and the second target region as the detected target region set.

[0011] In some embodiments, the method further includes:

[0012] Calculate multiple gradient values for each pixel point in the image to be detected;

[0013] Perform edge detection on the image to be detected based on the multiple gradient values of each pixel point in the image to be detected, and obtain a third target region of the image to be detected;

[0014] The step of determining the union of the first target region and the second target region as the detected target region set includes:

[0015] Determine the union of the first target region, the second target region, and the third target region as the target region set.

[0016] In some embodiments, the step of detecting the image to be detected based on a predefined plurality of anchor boxes to obtain a first target region of the image to be detected includes:

[0017] Input the image to be detected into a pre-trained first detection model;

[0018] In the first detection model, use a multi-layer convolutional neural network to extract a first feature map of the image to be detected;

[0019] Generate a plurality of anchor boxes on the first feature map based on a predefined plurality of grid cells and aspect ratios, and the center point of the anchor box is the center point of the corresponding grid cell;

[0020] For each anchor box, detect a target object region in a part of the first feature map included in the anchor box;

[0021] Adjust the position and size of the anchor box based on the anchor box and the corresponding target object region to obtain an adjusted anchor box;

[0022] Calculate the confidence of the adjusted anchor box based on the adjusted anchor box and the corresponding target object region;

[0023] For each group of overlapping anchor boxes, retain the anchor box with the highest confidence in a group of overlapping anchor boxes to obtain at least one detection box;

[0024] Determine the first target region based on all the detection boxes.

[0025] In some embodiments, the step of adjusting the position and size of the anchor box based on the anchor box and the corresponding target object region to obtain an adjusted anchor box includes:

[0026] Move the corresponding anchor box according to the center point of the target object area so that the center point of the anchor box coincides with the center point of the target object area;

[0027] Adjust the width and height of the anchor box based on the width and height of the target object area to obtain an adjusted anchor box.

[0028] In some embodiments, calculating the confidence of the adjusted anchor box based on the adjusted anchor box and the corresponding target object area includes:

[0029] Calculate the object existence probability of the adjusted anchor box based on the width and height of the adjusted anchor box and the width and height of the corresponding target object area;

[0030] Calculate the object category probability of the target object area;

[0031] Multiply the object existence probability and the object category probability to obtain the confidence of the adjusted anchor box.

[0032] In some embodiments, performing multiple convolutional operations and deconvolutional operations on the image to be detected to classify the objects in the image to be detected to obtain a second target area of the image to be detected includes:

[0033] Input the image to be detected into a pre-trained second detection model;

[0034] In the second detection model, perform multiple encoding operations including convolutional operations and decoding operations including deconvolutional operations on the image to be detected, so as to extract a second feature map of the image to be detected;

[0035] Perform a convolutional operation on the second feature map to obtain a class vector, and the class vector is used to represent the class of the object to which each pixel point in the image to be detected belongs;

[0036] Determine the second target area of the image to be detected based on the class vector.

[0037] In some embodiments, in the second detection model, performing multiple encoding operations including convolutional operations and decoding operations including deconvolutional operations on the image to be detected, so as to extract a second feature map of the image to be detected includes:

[0038] Input the image to be detected into the encoding layer of the second detection model, and use the encoding layer to perform multiple encoding operations including convolutional operations on the image to be detected to obtain a first feature vector, where the encoding layer includes a plurality of cascaded encoders, and each encoder is used to perform one encoding operation;

[0039] Input the first feature vector into the decoding layer of the second detection model, and use the decoding layer to perform multiple decoding operations including deconvolution operations on the first feature vector, so as to extract the second feature map of the image to be detected. Among them, the decoding layer includes a plurality of cascaded decoders, each decoder is connected to the encoder of the corresponding level, and each decoder is used to perform a decoding operation based on the feature vector input to the decoder and the feature vector output by the encoder of the corresponding level obtained.

[0040] In some embodiments, the method further includes:

[0041] Obtain an alignment degree detection image;

[0042] Extract a plurality of battery cell feature regions based on the alignment degree detection image;

[0043] Generate a calculation line segment based on each battery cell feature region, so as to obtain a plurality of calculation line segments, and determine the coordinates of the two endpoints of each calculation line segment. Among them, the first endpoint of the calculation line segment is located in the cathode region of the battery cell, and the second endpoint is located in the anode region of the battery cell;

[0044] Calculate the alignment degree of the battery cell based on the maximum coordinate value and the minimum coordinate value of all the first endpoints on the coordinate axis parallel to the calculation line segment, and the maximum coordinate value and the minimum coordinate value of all the second endpoints on the coordinate axis parallel to the calculation line segment.

[0045] In a second aspect, an embodiment of the present application provides an electronic device, and the electronic device includes:

[0046] At least one processor; and,

[0047] A memory communicatively connected to the at least one processor; wherein,

[0048] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the battery cell detection method as described in the first aspect.

[0049] In a third aspect, an embodiment of the present application provides a battery cell detection device, and the battery cell detection device includes a conveying device, an X-ray image acquisition device and the electronic device as described in the first aspect. The conveying device is used to convey the battery cell to the position where the X-ray image acquisition device is located, and the X-ray image acquisition device is used to acquire the image to be detected of the battery cell and send the image to be detected to the electronic device for processing, and the image to be detected is an X-ray image.

[0050] The present application provides a method for detecting an electric core, an electronic device, and an electric core detection device. By detecting a to-be-detected image based on a plurality of predefined anchor boxes, a first target region of the to-be-detected image is obtained; the to-be-detected image is subjected to multiple convolutional operations and deconvolution operations to classify the objects in the to-be-detected image, obtaining a second target region of the to-be-detected image, and the union of the first target region and the second target region is determined as the detected target region set, which can accurately detect all foreign objects in the electric core, thereby improving the comprehensiveness and accuracy of detection and avoiding problems of performance degradation and safety hazards of the electric core caused by foreign objects. Description of the Drawings

[0051] Figure 1 It is a schematic flowchart of the first implementation manner of the electric core detection method provided by an embodiment of the present application.

[0052] Figure 2 It is a schematic structural diagram of the first detection model provided by an embodiment of the present application.

[0053] Figure 3 It is a schematic structural diagram of the second detection model provided by an embodiment of the present application.

[0054] Figure 4 It is a schematic flowchart of the second implementation manner of the electric core detection method provided by an embodiment of the present application.

[0055] Figure 5 It is a schematic flowchart of the third implementation manner of the electric core detection method provided by an embodiment of the present application.

[0056] Figure 6 It is a schematic flowchart of the fourth implementation manner of the electric core detection method provided by an embodiment of the present application.

[0057] Figure 7 It is a result diagram of the target region set of the electric core provided by an embodiment of the present application.

[0058] Figure 8 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application.

[0059] Figure 9 It is a structural block diagram of a computer-readable storage medium provided by an embodiment of the present application.

[0060] Figure 10 It is a schematic structural diagram of the electric core detection device provided by an embodiment of the present application. Detailed Embodiments

[0061] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0062] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0063] The present application provides a battery cell detection method, an electronic device, and a battery cell detection device. By detecting a to-be-detected image based on a plurality of predefined anchor boxes, a first target area of the to-be-detected image is obtained; the to-be-detected image is subjected to multiple convolution operations and deconvolution operations to classify the objects in the to-be-detected image, obtaining a second target area of the to-be-detected image, and the union of the first target area and the second target area is determined as the detected target area set, which can accurately detect all foreign objects in the battery cell, thereby improving the comprehensiveness and accuracy of the detection and avoiding problems of performance degradation and safety hazards of the battery cell caused by foreign objects.

[0064] The battery cell detection method provided by the present application will be specifically described below with reference to the accompanying drawings.

[0065] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of the first implementation manner of the battery cell detection method provided by the embodiments of the present application. As Figure 1 shown, the battery cell detection method includes: step S100 to step S400.

[0066] Step S100: Obtain a to-be-detected image of a battery cell.

[0067] In some embodiments, the to-be-detected image of the battery cell is an X-ray image obtained by scanning the battery cell with an X-ray image acquisition device.

[0068] Step S200: Detect the to-be-detected image based on a plurality of predefined anchor boxes to obtain a first target area of the to-be-detected image.

[0069] In some embodiments, step S200 includes step S210 to step S280.

[0070] Step S210: Input the to-be-detected image into a pre-trained first detection model.

[0071] Please refer to Figure 2 , Figure 2 which is a schematic structural diagram of the first detection model provided by an embodiment of the present application. As Figure 2 shown, in some embodiments, the first detection model 10 includes a feature extraction module 101, a feature fusion module 102, a prediction module 103, and a result output module 104. Figure 2 The arrow direction in

[0072] represents the information flow direction. The output of the feature extraction module 101 is the input of the feature fusion module 102, the output of the feature fusion module 102 is the input of the prediction module 103, and the output of the prediction module 103 is the input of the result output module 104.

[0073] In some embodiments, the first detection model further includes a data preprocessing module. The data preprocessing module is used to preprocess the input image to be detected.

[0074] In some embodiments, the feature extraction module is used to extract features from the image to be detected, obtaining a third feature map. Among them, there are multiple third feature maps, and the scales (i.e., resolutions) of the multiple third feature maps are different.

[0075] Optionally, the feature extraction module is used to extract features from the preprocessed image to be detected output by the data preprocessing module, obtaining a third feature map.

[0076] Optionally, the feature extraction module includes multiple cascaded sub-feature extraction modules.

[0077] In some embodiments, the feature extraction module includes a CSPDarknet (Cross Stage Partial Network Darknet) module. Darknet refers to the original Darknet network architecture. In the CSPDarknet module, the feature map is divided into two parts. One part directly undergoes a downsampling operation, and the other part is concatenated with the feature map after the downsampling operation in a later stage. In this way, the computational amount can be reduced.

[0078] In some embodiments, the feature fusion module is used to fuse the third feature maps output by the feature extraction module, obtaining a first feature map. Among them, there are multiple first feature maps, and the scales of the multiple first feature maps are different.

[0079] Optionally, the feature fusion module includes multiple cascaded sub-feature fusion modules.

[0080] In some embodiments, the PANet (Pyramid Attention Network) structure is adopted in the feature fusion module to fuse high-resolution feature maps with low-resolution feature maps. In this way, the detection ability for small targets can be improved.

[0081] In some embodiments, the prediction module is used to generate a plurality of anchor boxes on each first feature map output by the feature fusion module based on a plurality of predefined grid cells and aspect ratios, and output the anchor box generation result to the result output module.

[0082] In some embodiments, the result output module is used to determine the first target region based on the anchor box generation result of the prediction module.

[0083] In some embodiments, steps S220 to S280 are all executed by the corresponding modules in the first detection model.

[0084] In some embodiments, before step S220, step S200 further includes: preprocessing the image to be detected to obtain the preprocessed image to be detected.

[0085] In some embodiments, the image to be detected is input into the data preprocessing module, and the data preprocessing module is used to preprocess the image to be detected to obtain the preprocessed image to be detected.

[0086] Optionally, the image to be detected is scaled to a preset image size while maintaining the aspect ratio of the image, the pixel values of all pixel points are normalized within the numerical range of 0 to 1, and then the image to be detected after the above preprocessing operations is converted into a preset tensor format to obtain the preprocessed image to be detected. At this time, the image to be detected in the subsequent steps all refers to the preprocessed image to be detected.

[0087] In some embodiments, the first detection model is constructed using the PyTorch framework, so the image to be detected is converted into a preset PyTorch tensor format. The PyTorch framework is an open-source deep learning framework widely used in the field of deep learning. In PyTorch, a tensor is a class of multi-dimensional matrices that can store multi-dimensional data, such as scalars, vectors, matrices, or higher-dimensional arrays, etc. That is, the preset tensor format includes formats such as scalars, vectors, matrices, or higher-dimensional arrays.

[0088] In some embodiments, after obtaining the preprocessed image to be detected, all the calculation processes can be transferred to the GPU for accelerated execution using the GPU.

[0089] In some embodiments, when the obtained image to be detected has been preprocessed, there is no need to preprocess the image to be detected again.

[0090] Step S220: In the first detection model, a multi-layer convolutional neural network is used to extract the first feature map of the image to be detected.

[0091] In some embodiments, the feature extraction module and the feature fusion module of the first detection model constitute a multi-layer convolutional neural network, and the feature extraction module and the feature fusion module of the first detection model are used to extract the first feature map of the image to be detected.

[0092] Optionally, the feature extraction module is used to extract features from the pre-processed image to be detected output by the data pre-processing module to obtain a third feature map. Among them, there are multiple third feature maps, and the scales (i.e., resolutions) of the multiple third feature maps are different.

[0093] Optionally, the feature fusion module is used to fuse the third feature maps output by the feature extraction module to obtain a first feature map. Among them, there are multiple first feature maps, and the scales of the multiple first feature maps are different.

[0094] Step S230: Based on a predefined plurality of grid cells and aspect ratios, a plurality of anchor boxes are generated on the first feature map, and the center point of the anchor box is the center point of the corresponding grid cell.

[0095] In some embodiments, multiple first feature maps are input into the prediction module, and the prediction module is used to generate a plurality of anchor boxes on each first feature map output by the feature fusion module based on a predefined plurality of grid cells and aspect ratios, and output the anchor box generation result to the result output module.

[0096] Optionally, the anchor box generation result includes the coordinates, width, height, corresponding object category, and confidence of the center point of each anchor box.

[0097] In some embodiments, the prediction module is used to execute Step S230 to Step S260.

[0098] In some embodiments, for each scale of the first feature map, the first feature map of this scale is pre-divided into a plurality of grid cells, and a plurality of anchor boxes are predefined in each grid cell. The center points of the plurality of anchor boxes are all the center points of the grid cell, and the aspect ratios, widths, and heights of the plurality of anchor boxes are different from each other. Therefore, each scale of the first feature map corresponds to a set of predefined plurality of grid cells and aspect ratios of the anchor boxes.

[0099] In some embodiments, a corresponding set of predefined plurality of grid cells and aspect ratios of the anchor boxes are selected based on the scale of the first feature map to generate a plurality of anchor boxes on the first feature map. Specifically, a plurality of anchor boxes are generated on the first feature map according to the predefined plurality of grid cells and the aspect ratios, widths, and heights of the plurality of anchor boxes corresponding to each grid cell.

[0100] Step S240: For each anchor box, detect the target object region in the partial first feature map included in the anchor box.

[0101] Optionally, the target object is a foreign object in the battery cell, and the target object region is the display region of the foreign object in the battery cell.

[0102] In some embodiments, the target object region recognized through feature extraction is already included in the first feature map. In step S240, detect the partial target object region included within the anchor box. In this case, the anchor box may only include a partial target object region and not the entire target object region.

[0103] Step S250: Based on the anchor box and the corresponding target object region, adjust the position and size of the anchor box to obtain an adjusted anchor box.

[0104] In some embodiments, step S250 includes steps S251 to S252.

[0105] Step S251: Move the corresponding anchor box according to the center point of the target object region so that the center point of the anchor box coincides with the center point of the target object region.

[0106] Optionally, the anchor box is rectangular, and the center point of the anchor box is the geometric center point of the rectangle.

[0107] Optionally, generate the smallest ground truth box containing the target object region according to the width and height of the target object region, and the center point of the target object region is the geometric center point of the ground truth box.

[0108] Step S252: Based on the width and height of the target object region, adjust the width and height of the anchor box to obtain an adjusted anchor box.

[0109] In some embodiments, calculate a width scaling factor based on the width of the target object region and the width of the anchor box, calculate a height scaling factor based on the height of the target object region and the height of the anchor box, and then adjust the width of the anchor box based on the width scaling factor and adjust the height of the anchor box based on the height scaling factor.

[0110] In some embodiments, the adjusted anchor box is not larger than the anchor box before adjustment, that is, the width of the adjusted anchor box is not larger than the width of the anchor box before adjustment, and the height of the adjusted anchor box is not larger than the height of the anchor box before adjustment.

[0111] Optionally, the formula for the width of the adjusted anchor box is: b w = a w e tw where b w represents the width of the adjusted anchor box, and aw w represents the width of the anchor box before adjustment, e represents the base of the natural logarithm, and tw represents the width scaling factor.

[0112] Optionally, the formula for calculating the height of the adjusted anchor box is: b h = a h e th , where b h represents the height of the adjusted anchor box, a h represents the height of the anchor box before adjustment, e represents the base of the natural logarithm, and th represents the height scaling factor.

[0113] Step S260: Calculate the confidence of the adjusted anchor box based on the adjusted anchor box and the corresponding target object region.

[0114] The target object region corresponding to the adjusted anchor box is a target object region where the center point of the anchor box is located.

[0115] In some embodiments, step S260 includes step S261 to step S263.

[0116] Step S261: Calculate the object existence probability of the adjusted anchor box based on the width and height of the adjusted anchor box, and the width and height of the corresponding target object region.

[0117] In some embodiments, a minimum true box containing the target object region is generated according to the width and height of the target object region, and the similarity between the adjusted anchor box and the true box corresponding to the target object region is calculated based on the width and height of the adjusted anchor box and the width and height of the true box corresponding to the target object region.

[0118] Optionally, calculate the intersection over union (IoU) of the adjusted anchor box and the corresponding true box and use it as the similarity between the adjusted anchor box and the corresponding true box.

[0119] Optionally, the intersection over union (IoU) of the adjusted anchor box and the corresponding true box refers to the ratio of the area of the intersection region of the adjusted anchor box and the corresponding true box to the area of the union region. As described above, the anchor box may only include a partial target object region and not the entire target object region. Therefore, the value range of the intersection over union (IoU) of the adjusted anchor box and the corresponding true box is from 0 to 1.

[0120] In some embodiments, when the similarity between the adjusted anchor box and the corresponding true box is greater than the similarity threshold, set the object existence probability of the adjusted anchor box to 1, otherwise set the object existence probability of the adjusted anchor box to 0.

[0121] Optionally, the similarity threshold is 0.5.

[0122] In some embodiments, the similarity between the adjusted anchor box and the corresponding ground truth box is determined as the object existence probability of the adjusted anchor box.

[0123] Step S262: Calculate the object category probability of the target object region.

[0124] Among them, the object category probability represents the probability that the object in the target object region belongs to a certain object category.

[0125] Optionally, the object categories include battery cells and foreign objects.

[0126] In some embodiments, a convolutional neural network is used to calculate the object category probability of the target object region.

[0127] In some embodiments, when the feature fusion module outputs the first feature map, it simultaneously identifies the target object region in the first feature map and calculates the object category probability of the target object region.

[0128] Step S263: Multiply the object existence probability and the object category probability to obtain the confidence of the adjusted anchor box.

[0129] Optionally, the anchor box generation result includes the coordinates of the center point, width, height, corresponding object category, and confidence of each adjusted anchor box. The object category corresponding to the anchor box is the object category of the target object region included in the anchor box.

[0130] In some embodiments, a result output module is used to determine the first target region based on the anchor box generation result of the prediction module.

[0131] Step S270: For each group of overlapping anchor boxes, retain the anchor box with the highest confidence in a group of overlapping anchor boxes to obtain at least one detection box.

[0132] In some embodiments, for each group of overlapping anchor boxes, a result output module is used to retain the anchor box with the highest confidence in a group of overlapping anchor boxes to obtain at least one detection box.

[0133] In some embodiments, when an anchor box does not overlap with other anchor boxes, the anchor box is also a detection box.

[0134] Step S280: Determine the first target region based on all detection boxes.

[0135] In some embodiments, a result output module is used to determine the first target region based on all detection boxes.

[0136] Optionally, the first target region is the union of the partial regions in the first feature map included in all detection boxes.

[0137] In some embodiments, the object category, confidence level, coordinates of the center point, width, and height of each detection box are obtained simultaneously.

[0138] In some embodiments, the first target area is the display area of fine foreign objects in the battery cell. Since the gaskets on both sides of the battery cell are slightly thicker, the image to be detected will include two black edge areas and a large white area in the middle. By using the method of determining the first target area with the first detection model, the display area of fine foreign objects can be located through anchor boxes according to different areas of the image to be detected, so as to accurately detect the fine foreign objects in the battery cell.

[0139] Step S300: Perform multiple convolution operations and deconvolution operations on the image to be detected to classify the objects in the image to be detected, and obtain the second target area of the image to be detected.

[0140] Among them, the objects in the image to be detected include the battery cell and foreign objects.

[0141] In some embodiments, step S300 includes step S310 to step S340.

[0142] Step S310: Input the image to be detected into a pre-trained second detection model.

[0143] Please refer to Figure 3 , Figure 3 which is a schematic structural diagram of the second detection model provided by the embodiments of the present application. Figure 3 The arrow direction in Figure 3 indicates the information flow direction. As

[0144] In some embodiments, the second detection model is a neural network model. For example, the second detection model is a U-net model or an improved U-net model.

[0145] Step S320: In the second detection model, perform multiple encoding operations including convolution operations and decoding operations including deconvolution operations on the image to be detected, so as to extract the second feature map of the image to be detected.

[0146] In some embodiments, step S320 includes step S321 to step S322.

[0147] Step S321: Input the image to be detected into the encoding layer of the second detection model, and use the encoding layer to perform multiple encoding operations including convolution operations on the image to be detected to obtain the first feature vector.

[0148] As Figure 3As shown, in some embodiments, the encoding layer 21 includes a plurality of cascaded encoders, and each encoder is used to perform an encoding operation once.

[0149] Preferably, the number of encoders is 4.

[0150] It is defined that in the encoding layer, the encoder that obtains the image to be detected is the first cascaded encoder. As Figure 3 shown, exemplarily, encoder 211 is the first cascaded encoder, encoder 212 is the second cascaded encoder, encoder 213 is the third cascaded encoder, and encoder 214 is the fourth cascaded encoder.

[0151] In some embodiments, an encoder performs multiple convolution operations on the obtained image to be detected or the third feature vector output by the previous-level encoder to obtain a third feature vector. When this encoder is not the last-level encoder, the third feature vector is input to the next-level encoder.

[0152] As Figure 3 shown, in some embodiments, the encoding layer 21 further includes an FPA (feature pyramid attention, FPA feature pyramid) module 215. The FPA module 215 is used to perform dilated convolution operations, global average pooling operations, and upsampling operations on the third feature vector output by the last-level encoder in sequence to obtain a fourth feature vector, and then perform feature concatenation on the fourth feature vector and the third feature vectors output by all encoders except the last-level encoder to obtain a first feature vector.

[0153] In some other embodiments, when the encoder is the last-level encoder, the third feature vector is directly output as the first feature vector.

[0154] Step S322: Input the first feature vector into the decoding layer of the second detection model, and use the decoding layer to perform multiple decoding operations including deconvolution operations on the first feature vector, so as to extract the second feature map of the image to be detected.

[0155] Among them, the decoding layer includes a plurality of cascaded decoders, each decoder is connected to the encoder of the corresponding level, and each decoder is used to perform a decoding operation based on the feature vector input to the decoder and the feature vector output by the encoder of the corresponding level obtained.

[0156] It is defined that in the decoding layer, the decoder that outputs the second feature vector is the first cascaded decoder. As Figure 3 shown, exemplarily, decoder 221 is the first cascaded decoder, decoder 222 is the second cascaded decoder, decoder 223 is the third cascaded decoder, and decoder 224 is the fourth cascaded decoder.

[0157] As Figure 3 shown, optionally, when the decoder 224 is the last-stage decoder of the decoding layer 22, the decoder 224 performs a decoding operation only once based on the first feature vector output by the FPA module 215.

[0158] Optionally, the encoder of the corresponding stage of the decoder includes multiple encoders. The encoder of the corresponding stage of the decoder refers to the encoder whose cascading order in the encoding layer is not greater than the cascading order of the decoder in the decoding layer.

[0159] As Figure 3 shown, exemplarily, if the decoder 223 is the third cascaded decoder in the decoding layer 22, the encoders of the corresponding stage of the decoder 223 include the encoder 221, the encoder 222, and the encoder 223. The decoder 223 performs a decoding operation once based on the feature vector input to the decoder 223 and the feature vectors output by the encoder 221, the encoder 222, and the encoder 223.

[0160] In some embodiments, a decoder performs multiple deconvolution operations on the obtained first feature vector to obtain a fifth feature vector, and then concatenates the fifth feature vector with the third feature vectors output by all the encoders of the corresponding stage to obtain a sixth feature vector. When the decoder is not the last-stage encoder, the sixth feature vector is input to the next-stage encoder. When the decoder is the last-stage encoder, the sixth feature vector is output as the second feature map. It can be understood that a vector can be represented as an image, and an image can also be represented in the form of a vector.

[0161] Step S330: Perform a convolution operation on the second feature map to obtain a class vector.

[0162] Wherein, the class vector is used to represent the class of the object to which each pixel point in the image to be detected belongs.

[0163] As Figure 3 shown, in some embodiments, the result output layer 23 performs a convolution operation on the second feature map output by the first cascaded decoder 221 to obtain a class vector.

[0164] Step S340: Determine the second target region of the image to be detected based on the class vector.

[0165] In some embodiments, the union of all pixel points whose class of the object is a foreign object is determined as the second target region.

[0166] In some embodiments, the second target region is the display region of a relatively large foreign object in the battery cell. By using the second detection model to determine the second target region, it is possible to locate and segment the display region of the relatively large foreign object, thereby accurately detecting the relatively large foreign object in the battery cell.

[0167] Step S400: Determine the union of the first target region and the second target region as the detected target region set.

[0168] Please refer to Figure 4 , Figure 4 which is a schematic flowchart of the second embodiment of the battery cell detection method provided by the embodiments of the present application. As Figure 4 shown, the battery cell detection method further includes step S500.

[0169] Step S500: Perform edge detection on the image to be detected to obtain the third target region of the image to be detected.

[0170] In some embodiments, step S500 includes step S510 to step S520.

[0171] Step S510: Calculate multiple gradient values of each pixel point in the image to be detected.

[0172] In some embodiments, first perform a Gaussian blur operation on the image to be detected, and then perform a convolution operation on the image to be detected after the Gaussian blur operation to obtain the gradient values in the horizontal and vertical directions of each pixel point.

[0173] Optionally, in the convolution operation, use a Sobel convolution kernel to perform a convolution operation on the image to be detected.

[0174] Step S520: Based on the multiple gradient values of each pixel point in the image to be detected, perform edge detection on the image to be detected to obtain the third target region of the image to be detected.

[0175] In some embodiments, based on a preset gradient threshold and the multiple gradient values of each pixel point in the image to be detected, perform edge detection on the image to be detected to obtain the third target region of the image to be detected. Specifically, when the multiple gradient values of a pixel point are all less than the gradient threshold, determine the pixel point as a non-edge pixel; when the multiple gradient values of a pixel point are not all less than the gradient threshold, determine the pixel point as an edge pixel. Finally, based on all edge pixels, the region boundary of the third target region can be determined, and then the third target region can be determined.

[0176] In some embodiments, the third target area is a display area for foreign objects that are difficult to detect on the surface and edges of the battery cell. By performing edge detection on the image to be detected based on a preset gradient threshold and multiple gradient values of each pixel point in the image to be detected, non-edge pixels can be filtered out, so as to accurately detect foreign objects that are difficult to detect on the surface and edges of the battery cell.

[0177] In some embodiments, when the battery cell detection method further includes step S500, step S400 includes: determining the union of the first target area, the second target area, and the third target area as the target area set.

[0178] By determining the union of the first target area, the second target area, and the third target area as the target area set, all foreign objects in the battery cell can be further accurately detected, thereby further improving the comprehensiveness and accuracy of the detection, and avoiding problems such as performance degradation and safety hazards of the battery cell caused by foreign objects.

[0179] Please refer to Figure 5 , Figure 5 which is a schematic flowchart of the 3rd embodiment of the battery cell detection method provided by the embodiments of the present application. As Figure 5 shown, in some embodiments, the battery cell detection method further includes step S600.

[0180] Step S600: Perform alignment detection on the battery cell to obtain the alignment of the battery cell.

[0181] In some embodiments, the battery cell is a cuboid.

[0182] As Figure 5 shown, in some embodiments, before step S100, step S600 is executed to perform alignment detection on the first pair of adjacent corner points of the battery cell to obtain the first alignment of the battery cell. After step S400 or step S500, step S600 is executed to perform alignment detection on the second pair of adjacent corner points of the battery cell to obtain the second alignment of the battery cell.

[0183] In some embodiments, step S600 includes steps S610 to S640.

[0184] Step S610: Obtain an alignment detection image.

[0185] In some embodiments, before step S100, an alignment detection image of the first pair of adjacent corner points of the battery cell is obtained.

[0186] In some embodiments, after step S400 or step S500, an alignment detection image of the second pair of adjacent corner points of the battery cell is obtained. Among them, the two corner points included in the second pair of adjacent corner points do not overlap with the two corner points included in the first pair of adjacent corner points.

[0187] Step S620: Extract a plurality of battery cell feature regions based on the alignment detection image.

[0188] In some embodiments, the alignment detection image is preprocessed first. Optionally, the alignment detection image is filtered with a filter to reduce image noise, and histogram equalization operation and contrast adjustment operation are performed on the alignment detection image to enhance the image. Then, an image segmentation algorithm is used to detect the edge features in the alignment detection image, and the battery cell feature regions are separated from the non-battery cell feature regions according to the edge features, and the battery cell feature regions are retained.

[0189] Optionally, the filter includes a Gaussian filter, a median filter, etc.

[0190] Optionally, the image segmentation algorithm includes a Canny segmentation algorithm, a Soble segmentation algorithm, etc.

[0191] Step S630: Generate a calculation line segment based on each battery cell feature region, so as to obtain a plurality of calculation line segments, and determine the coordinates of the two endpoints of each calculation line segment.

[0192] Among them, the first endpoint of the calculation line segment is located in the cathode region of the battery cell, and the second endpoint is located in the anode region of the battery cell.

[0193] In some embodiments, the battery cell feature region is rectangular or approximately rectangular, and a calculation line segment is fitted based on the two long sides of the battery cell feature region.

[0194] Step S640: Calculate the alignment degree of the battery cell based on the maximum coordinate value and the minimum coordinate value of all the first endpoints on the coordinate axis parallel to the calculation line segment, and the maximum coordinate value and the minimum coordinate value of all the second endpoints on the coordinate axis parallel to the calculation line segment.

[0195] In some embodiments, the maximum coordinate value and the minimum coordinate value of the first endpoint on the coordinate axis parallel to the calculation line segment are subtracted to obtain the alignment degree of the cathode region corresponding to a pair of adjacent corner points.

[0196] In some embodiments, the maximum coordinate value and the minimum coordinate value of the second endpoint on the coordinate axis parallel to the calculation line segment are subtracted to obtain the alignment degree of the anode region corresponding to a pair of adjacent corner points.

[0197] In some embodiments, the first alignment degree includes the first cathode region alignment degree and the first anode region alignment degree. The second alignment degree includes the second cathode region alignment degree and the second anode region alignment degree.

[0198] Please refer to Figure 6 , Figure 6 which is a schematic flowchart of the 4th embodiment of the battery cell detection method provided by the embodiments of the present application. As Figure 6 shown, in some embodiments, the battery cell detection method further includes step S700.

[0199] Step S700: Output a comprehensive detection result of the battery cell based on the alignment degree of the battery cell, the target area set, and a preset judgment condition.

[0200] In some embodiments, when the value of any alignment degree of the battery cell is not greater than a preset alignment degree threshold and the target area set is empty, it is determined that the comprehensive detection result of the battery cell is qualified.

[0201] In some embodiments, when the value of any alignment degree of the battery cell is greater than the preset alignment degree threshold, or the target area set is not empty, it is determined that the comprehensive detection result of the battery cell is unqualified.

[0202] In some embodiments, when the number of target areas in the target area set is less than a judgment number threshold, the steps in the above method are also re - used to re - detect the battery cell.

[0203] In some embodiments, when the number of target areas in the target area set is still less than the judgment number threshold in the comprehensive detection result output during the re - detection of the battery cell, it is determined that the comprehensive detection result of the battery cell is unqualified.

[0204] Please refer to Figure 7 , Figure 7 which is a result diagram of the target area set of the battery cell provided by the embodiments of the present application. As Figure 7 shown, in some embodiments, in the target area set result diagram P, when the target area set is not empty, the target areas in the target area set are also counted and numbered, and a display box A containing each target area is generated and displayed.

[0205] Optionally, the number of the target area is also displayed beside the display box A.

[0206] In summary, the battery cell detection method provided by the embodiments of the present application has the following advantages:

[0207] 1. By detecting the image to be detected based on a predefined plurality of anchor boxes, a first target region of the image to be detected is obtained; the image to be detected is subjected to multiple convolutional operations and deconvolution operations to classify the objects in the image to be detected, obtaining a second target region of the image to be detected, and the union of the first target region and the second target region is determined as the detected target region set, which can accurately detect all foreign objects in the battery cell, thereby improving the comprehensiveness and accuracy of detection and avoiding problems such as performance degradation and safety hazards of the battery cell caused by foreign objects.

[0208] 2. By using the method of determining the first target region with the first detection model, the display region of small foreign objects can be located through anchor boxes according to the regions of different images to be detected, so as to accurately detect small foreign objects in the battery cell.

[0209] 3. By using the method of determining the second target region with the second detection model, the display region of large foreign objects can be located and segmented, so as to accurately detect large foreign objects in the battery cell.

[0210] 4. By performing edge detection on the image to be detected based on a preset gradient threshold and multiple gradient values of each pixel point in the image to be detected, non-edge pixels can be filtered out, so as to accurately detect foreign objects on the surface and edges of the battery cell that are difficult to detect.

[0211] 5. By

[0212] Please refer to Figure 8 , Figure 8 which is a schematic structural diagram of an electronic device provided by an embodiment of the present application. As Figure 8 shown, the electronic device 400 includes: one or more processors 410 and a memory 420, Figure 8 taking one processor 410 as an example in

[0213] In some embodiments, the processor 410 and the memory 420 may be connected by a bus or other means, Figure 8 taking connection by bus as an example in

[0214] In some embodiments, the processor 410 is configured to obtain an image to be detected of the battery cell; detect the image to be detected based on a predefined plurality of anchor boxes to obtain a first target region of the image to be detected; perform multiple convolutional operations and deconvolution operations on the image to be detected to classify the objects in the image to be detected, obtaining a second target region of the image to be detected, where the objects in the image to be detected include the battery cell and foreign objects; and determine the union of the first target region and the second target region as the detected target region set.

[0215] In some embodiments, the memory 420 serves as a non-volatile computer-readable storage medium and can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules of the cell detection method in the embodiments of the present application. The processor 410 executes various functional applications and data processing of the electronic device 400 by running the non-volatile software programs, instructions, and modules stored in the memory 420, that is, implements the cell detection method in the above method embodiments.

[0216] In some embodiments, the memory 420 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the electronic device 400, etc. In addition, the memory 420 may include high-speed random access memory and may also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 420 optionally includes a memory remotely set relative to the processor 410, and these remote memories can be connected to the controller through a network. Examples of the above networks include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0217] In some embodiments, one or more modules are stored in the memory 420 and, when executed by one or more processors 410, implement the cell detection method in any of the above method embodiments. For example, execute the Figure 1 method steps S100 to step S400 described above.

[0218] Please refer to Figure 9 , Figure 9 , which is a structural block diagram of a computer-readable storage medium provided by an embodiment of the present application. Program code 510 is stored in the computer-readable storage medium 500, and the program code 510 can be called by a processor to execute the cell detection method described in the above method embodiments.

[0219] The computer-readable storage medium 500 can be an electronic memory such as a flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, a hard disk, or a ROM. Optionally, the computer-readable storage medium includes a non-transitory computer-readable storage medium. The computer-readable storage medium 500 has a storage space for program code that executes any method step in the above cell detection method. These program codes can be read out from or written into one or more computer program products. The program codes can be compressed in an appropriate form, for example.

[0220] In some embodiments, the embodiments of the present application further provide a computer program product, including a computer program, which when executed by a processor implements the above-mentioned battery cell detection method.

[0221] The present application further provides a battery cell detection device, which includes a conveying device, an X-ray image acquisition device and the electronic device as described above. The conveying device is used to convey the battery cell to the position where the X-ray image acquisition device is located. The X-ray image acquisition device is used to acquire the image to be detected of the battery cell and send the image to be detected to the electronic device for processing. Among them, the image to be detected is an X-ray image.

[0222] Please refer to Figure 10 , Figure 10 which is a schematic structural diagram of the battery cell detection device provided by the embodiments of the present application. As Figure 10 shown, in some embodiments, the conveying device of the battery cell detection device 3 includes a detection conveyor belt 31 and a retest conveyor belt 39. The detection conveyor belt 31 is used to convey the battery cell to be initially detected to the position where the X-ray image acquisition device is located. The retest conveyor belt 39 is used to convey the battery cell that needs to be retested to the position where the X-ray image acquisition device is located. The X-ray image acquisition device of the battery cell detection device 3 includes a first X-ray image acquisition device 35 and a second X-ray image acquisition device 36. The first X-ray image acquisition device 35 is used to acquire the image to be detected of the battery cell. The second X-ray image acquisition device 36 is used to acquire the alignment detection image of the battery cell. The electronic device 400 is used to execute the above-mentioned battery cell detection method.

[0223] In some embodiments, the first X-ray image acquisition device 35 includes 2 sets of X-ray image acquisition components.

[0224] In some embodiments, the second X-ray image acquisition device 36 includes 8 sets of X-ray image acquisition components.

[0225] As Figure 10 shown, in some embodiments, the battery cell detection device 3 further includes a retest manipulator 32, an artificial processing station 33, a code scanning device 34, an NG belt strip 37, an NG manipulator 38, a first blanking manipulator 41 and a second blanking manipulator 42. The retest manipulator 32 is used to take out the battery cell that needs to be retested and place it on the retest conveyor belt 39. The artificial processing station 33 is used to manually take out the unqualified battery cell at the artificial processing station 33. The code scanning device 34 is used to scan the unique identifier of the battery cell.

[0226] "NG" is used in manufacturing to represent "Not Good" or "defective product". In some embodiments, the NG manipulator 38 is used to take out the defective battery cells and place them on the NG belt 37. The first blanking manipulator 41 and the second blanking manipulator 42 are both used to take out the qualified battery cells.

[0227] Optionally, the first blanking manipulator 41 can take out 2 battery cells at a time.

[0228] Optionally, the second blanking manipulator 42 can take out 4 battery cells at a time.

[0229] In summary, the present application provides a battery cell detection method, an electronic device, and a battery cell detection device. The battery cell detection method includes: obtaining a to-be-detected image of a battery cell; detecting the to-be-detected image based on a plurality of predefined anchor boxes to obtain a first target region of the to-be-detected image; performing multiple convolution operations and deconvolution operations on the to-be-detected image to classify the objects in the to-be-detected image and obtain a second target region of the to-be-detected image, where the objects in the to-be-detected image include battery cells and foreign objects; and determining the union of the first target region and the second target region as the detected target region set. By detecting the to-be-detected image based on a plurality of predefined anchor boxes to obtain the first target region of the to-be-detected image, performing multiple convolution operations and deconvolution operations on the to-be-detected image to classify the objects in the to-be-detected image and obtain the second target region of the to-be-detected image, and determining the union of the first target region and the second target region as the detected target region set, the present application can accurately detect all foreign objects in the battery cell, thereby improving the comprehensiveness and accuracy of detection and avoiding problems such as performance degradation and safety hazards of the battery cell caused by foreign objects.

[0230] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A battery cell detection method, characterized in that: include: Acquire the image of the battery cell to be inspected; Detecting the image to be detected based on a plurality of predefined anchor frames to obtain a first target region of the image to be detected; Performing multiple convolution operations and deconvolution operations on the image to be detected to classify objects in the image to be detected to obtain a second target area of ​​the image to be detected, wherein the objects in the image to be detected include battery cells and foreign objects; A union of the first target area and the second target area is determined as a detected target area set.

2. The battery cell detection method according to claim 1, characterized in that: The method further comprises: Calculating multiple gradient values ​​for each pixel in the image to be detected; Performing edge detection on the image to be detected based on multiple gradient values ​​of each pixel in the image to be detected to obtain a third target area of ​​the image to be detected; The step of determining a union of the first target area and the second target area as a detected target area set includes: A union of the first target area, the second target area, and the third target area is determined as the target area set.

3. The battery cell detection method according to claim 1, characterized in that: The detecting the image to be detected based on a plurality of predefined anchor frames to obtain a first target area of ​​the image to be detected includes: Inputting the image to be detected into a pre-trained first detection model; In the first detection model, a multi-layer convolutional neural network is used to extract a first feature map of the image to be detected; Generate multiple anchor frames on the first feature map based on multiple predefined grid units and aspect ratios, where the center points of the anchor frames are the center points of the corresponding grid units; For each of the anchor frames, detecting a target object region in a portion of the first feature map included in the anchor frame; Adjusting the position and size of the anchor frame based on the anchor frame and the corresponding target object area to obtain an adjusted anchor frame; Calculating the confidence of the adjusted anchor frame based on the adjusted anchor frame and the corresponding target object area; For each set of overlapping anchor frames, retain the anchor frame with the highest confidence among the overlapping anchor frames to obtain at least one detection frame; The first target area is determined based on all the detection frames.

4. The battery cell detection method according to claim 3, characterized in that: The adjusting the position and size of the anchor frame based on the anchor frame and the corresponding target object area to obtain an adjusted anchor frame includes: Moving the corresponding anchor frame according to the center point of the target object area so that the center point of the anchor frame coincides with the center point of the target object area; The width and height of the anchor frame are adjusted based on the width and height of the target object area to obtain an adjusted anchor frame.

5. The battery cell detection method according to claim 3, characterized in that: The calculating the confidence of the adjusted anchor frame based on the adjusted anchor frame and the corresponding target object area includes: Calculating the object existence probability of the adjusted anchor frame based on the adjusted width and height of the anchor frame and the width and height of the corresponding target object area; Calculating the object category probability of the target object area; The object existence probability and the object category probability are multiplied to obtain the confidence of the adjusted anchor box.

6. The battery cell detection method according to claim 1, characterized in that: The performing multiple convolution operations and deconvolution operations on the image to be detected to classify the objects in the image to be detected and obtain the second target area of ​​the image to be detected includes: Inputting the image to be detected into a pre-trained second detection model; In the second detection model, a plurality of encoding operations including convolution operations and a decoding operation including deconvolution operations are performed on the image to be detected, so as to extract a second feature map of the image to be detected; Performing a convolution operation on the second feature map to obtain a category vector, where the category vector is used to represent the category of the object to which each pixel in the image to be detected belongs; A second object region of the image to be detected is determined based on the category vector.

7. The battery cell detection method according to claim 6, characterized in that: In the second detection model, performing a plurality of encoding operations including convolution operations and decoding operations including deconvolution operations on the image to be detected, thereby extracting a second feature map of the image to be detected, including: Inputting the image to be detected into the encoding layer of the second detection model, and using the encoding layer to perform multiple encoding operations including convolution operations on the image to be detected to obtain a first feature vector, wherein the encoding layer includes multiple cascaded encoders, each of which is used to perform one encoding operation; The first feature vector is input into the decoding layer of the second detection model, and the decoding layer is used to perform multiple decoding operations including deconvolution operations on the first feature vector, so as to extract the second feature map of the image to be detected, wherein the decoding layer includes multiple cascaded decoders, each of the decoders is connected to the encoder of the corresponding level, and each of the decoders is used to perform a decoding operation based on the feature vector input to the decoder and the feature vector output by the encoder of the corresponding level.

8. The battery cell detection method according to claim 1, characterized in that: The method further comprises: Acquire alignment detection image; Extracting a plurality of cell feature regions based on the alignment detection image; Generate a calculation line segment based on each of the battery cell characteristic regions, thereby obtaining a plurality of the calculation line segments, and determine the coordinates of two endpoints of each of the calculation line segments, wherein the first endpoint of the calculation line segment is located in the cathode region of the battery cell, and the second endpoint is located in the anode region of the battery cell; The alignment of the battery cell is calculated based on the maximum coordinate value and the minimum coordinate value of all the first endpoints on the coordinate axis parallel to the calculation line segment, and the maximum coordinate value and the minimum coordinate value of all the second endpoints on the coordinate axis parallel to the calculation line segment.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the battery cell detection method according to any one of claims 1 to 8.

10. A battery cell testing device, characterized in that: The battery cell detection equipment includes a transmission device, an X-ray image acquisition device and an electronic device as described in claim 9, wherein the transmission device is used to transmit the battery cell to the location of the X-ray image acquisition device, and the X-ray image acquisition device is used to acquire the image to be detected of the battery cell and send the image to be detected to the electronic device for processing, and the image to be detected is an X-ray image.

Images