Battery piece detection method, device, equipment and computer readable storage medium

Abstract

The present disclosure provides a battery piece detection method, device and equipment and a computer readable storage medium. The method comprises: obtaining a target picture, the target picture containing a photovoltaic module, the photovoltaic module containing a battery piece array formed by a plurality of battery pieces; segmenting the target picture to obtain a plurality of sub-pictures, wherein each sub-picture contains at least one battery piece; respectively detecting welding defects of the plurality of sub-pictures to obtain at least one candidate sub-picture with welding defects; and for the at least one candidate sub-picture, according to a screening condition corresponding to the candidate sub-picture, determining a welding defect in the candidate sub-picture that meets the screening condition as a target welding defect; wherein the screening conditions corresponding to at least two sub-pictures of the plurality of sub-pictures are different. The present disclosure can screen welding defects of battery pieces in different regions and different groups with higher accuracy.

Description

Technical Field

[0001] This disclosure relates to the field of photovoltaic technology, and in particular to a method, apparatus, device, and computer-readable storage medium for testing solar cells. Background Technology

[0002] Photovoltaic modules, also known as solar panels, are the core component of a solar power generation system. Their function is to convert solar energy into electrical energy, which is then stored in batteries or used to power loads.

[0003] The production and assembly process of solar panels generally includes multiple steps such as cell testing, cell stringing, welding inspection, laying, lamination, deburring, frame assembly, junction box welding, high-voltage testing, module testing, and appearance inspection. Among these, cell stringing typically involves connecting multiple solar cells in series with solder strips to form a solar cell string.

[0004] However, some incomplete solder joints can occur during the soldering process. These incomplete solder joints can cause the solder strip to detach from the solar cell within a short period, subsequently affecting the power output or even causing the photovoltaic module to fail. Currently, one method for detecting incomplete solder joints is through image recognition algorithms. However, because photovoltaic module designs vary, and the requirements for detecting incomplete solder joints are diverse, current detection methods can only perform overall identification of the photovoltaic module and cannot adapt to these variations. Summary of the Invention

[0005] This disclosure provides a method, apparatus, device, and computer-readable storage medium for testing solar cells, which can screen out welding defects in different regions and groups of solar cells with higher accuracy.

[0006] In a first aspect, this disclosure provides a method for testing solar cells, the method comprising:

[0007] Acquire a target image, wherein the target image contains a photovoltaic module, and the photovoltaic module comprises a cell array formed by multiple cells;

[0008] The target image is segmented to obtain multiple sub-images, wherein each sub-image contains at least one of the battery cells;

[0009] Welding defect detection is performed on the multiple sub-images respectively to obtain at least one candidate sub-image with welding defects;

[0010] For the at least one candidate sub-image, according to the filtering conditions corresponding to the candidate sub-image, the welding defects in the candidate sub-image that meet the filtering conditions are determined as target welding defects; wherein, at least two sub-images among the plurality of sub-images correspond to different filtering conditions.

[0011] Optionally, the plurality of sub-images are divided into at least two groups of sub-images, wherein the sub-images in the same group of sub-images have the same filtering conditions, and the sub-images in different groups of sub-images have different filtering conditions.

[0012] Optionally, the plurality of sub-images are divided into at least two groups according to the position of the battery cells in the sub-images in the battery cell array.

[0013] Optionally, solder strips are formed on the battery cell array, and the at least two groups are arranged in a direction parallel to the solder strips.

[0014] Optionally, among the plurality of sub-images, a first sub-image having battery cells located at the cut surface of the battery cell array and a second sub-image having battery cells located at the non-cut surface of the battery cell array are located in different groups.

[0015] Optionally, the battery cell array is provided with a busbar for electrically connecting at least two battery cells that are electrically connected to the busbar;

[0016] Among the multiple sub-images, the third sub-image, which has a battery cell that is in contact with the busbar or at a distance from the busbar that is no greater than a preset distance, is in a different group from the fourth sub-image, which has a battery cell that is at a distance from the busbar that is greater than the preset distance.

[0017] Optionally, the plurality of sub-images are divided into at least two groups of sub-images based on the chamfer direction of the battery cells in the sub-images.

[0018] Optionally, in the at least two sets of sub-images, each sub-image in at least one set of sub-images is adjacent to the next in the cell array.

[0019] Optionally, the step of detecting welding defects in the plurality of sub-images includes:

[0020] Welding defects are detected in the multiple sub-images using artificial intelligence algorithms.

[0021] Optionally, the filtering criteria include at least one of the following:

[0022] The ratio of the width of the welding defect to the width of the battery cell it belongs to is greater than a first threshold.

[0023] The ratio of the area of ​​the welding defect to the area of ​​the cell in which it is located is greater than the second threshold.

[0024] The grayscale value of the welding defect is within a preset grayscale range;

[0025] The welding defect is defined as a target welding defect whose confidence level is greater than a third threshold.

[0026] Secondly, this disclosure provides a testing device for solar cells, comprising:

[0027] An acquisition module is used to acquire a target image, wherein the target image contains a photovoltaic module, and the photovoltaic module contains a cell array formed by multiple cells.

[0028] A segmentation module is used to segment the target image to obtain multiple sub-images, wherein each sub-image contains at least one of the battery cells;

[0029] The detection module is used to detect welding defects in the multiple sub-images respectively, and obtain at least one candidate sub-image with welding defects;

[0030] A filtering module is used to identify welding defects in at least one candidate sub-image as target welding defects based on the filtering conditions corresponding to the candidate sub-image; wherein at least two of the multiple sub-images have different filtering conditions.

[0031] Thirdly, this disclosure provides a testing device for battery cells, including a memory and a processor, wherein the memory stores executable code, and when the executable code is processed by the processor, the processor can execute any of the methods described above.

[0032] Fourthly, this disclosure provides a computer-readable storage medium having executable code stored thereon, which, when executed by a processor, causes the processor to perform any of the methods described above.

[0033] In this embodiment of the disclosure, the image with the battery cell array is first segmented to obtain sub-images with at least one battery cell each. Then, welding defect detection is performed on different sub-images. After obtaining candidate sub-images with welding defects, different screening conditions can be conveniently used to screen the welding defects in different candidate sub-images to meet the different requirements for welding defects on battery cells in different regions and groups. This facilitates matching different demand scenarios and enables the screening of welding defects in different regions and groups with higher accuracy. Attached Figure Description

[0034] Figure 1 This is a flowchart of one embodiment of the battery cell testing method disclosed herein.

[0035] Figure 2 This is a partial structural schematic diagram of one embodiment of the battery cell array in this disclosure.

[0036] Figure 3This is a schematic diagram of another embodiment of the battery cell array in this disclosure.

[0037] Figure 4 This is a structural block diagram of one embodiment of the battery cell testing apparatus of this disclosure.

[0038] Figure 5 This is a structural block diagram of one embodiment of the battery cell testing device disclosed herein. Detailed Implementation

[0039] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present disclosure will be thorough and complete, and will fully convey the scope of the present disclosure to those skilled in the art.

[0040] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The singular forms “a,” “the,” and “the” as used in this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0041] It should be understood that although this disclosure may use terms such as "first," "second," and "third" to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this disclosure, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this disclosure, "a plurality of" means two or more, unless otherwise explicitly specified.

[0042] like Figure 1 As shown, Figure 1 This is a schematic flowchart of one embodiment of the battery cell testing method disclosed herein. The battery cell testing method 100 includes:

[0043] Step S101: Obtain a target image, wherein the target image contains a photovoltaic module, and the photovoltaic module contains a cell array formed by multiple cells.

[0044] A photovoltaic module comprises an array of solar cells. Optionally, these solar cells can be formed by string welding, single welding, splicing, or other methods. Specifically, each solar cell has at least two solder strips welded on it, and adjacent solar cells are connected in series through these solder strips. Figure 2 As shown, Figure 2 This is a partial structural schematic diagram of one embodiment of a solar cell array. The solar cell array includes multiple solar cells arranged in multiple rows and columns, such as solar cells 21, 23, and 24. Each solar cell has multiple parallel solder strips 22 soldered on it. One solder strip on each solar cell is connected to a solder strip on one of multiple other solar cells arranged in the same row to form a solder strip that spans across the solar cells in that row, so as to connect the solar cells in the row in series.

[0045] Step S102: Segment the target image to obtain multiple sub-images, wherein each sub-image contains at least one of the battery cells.

[0046] After obtaining the target image, individual battery cells can be extracted from it using image processing algorithms. For example, an image processing algorithm from an open-source computer vision library can be used to extract the individual battery cells, resulting in multiple sub-images. In one example, each sub-image contains only one battery cell. In another example, at least some sub-images contain at least two battery cells. The number of battery cells in different sub-images can be the same or different; this is not a restriction.

[0047] Optionally, while obtaining multiple sub-images, the position of each sub-image is also obtained, and the position of the solar cell in the solar cell array is determined based on the position of the sub-image, so as to subsequently determine the location of the detected welding defects.

[0048] Step S103: Perform welding defect detection on the multiple sub-images respectively to obtain at least one candidate sub-image with welding defects.

[0049] The power transmission function of a photovoltaic (PV) module is achieved by connecting multiple individual cells together with solder ribbons to form a single unit. However, improper control during the soldering process can lead to various welding defects, such as hot spots, incomplete soldering, over-soldering, detachment, and fragmentation. For example, incomplete soldering can cause the solder ribbon to detach from the cell within a short time, subsequently affecting the power output of the PV module or causing it to fail. Over-soldering can lead to crystal growth in the weld seam, reduced strength, or cell breakage, which in turn damages the electrodes inside the cell, directly affecting the power output of the PV module, thus reducing its lifespan or causing waste.

[0050] In one example, when detecting weld defects in various sub-images, one can detect only one type of weld defect or detect multiple weld defects simultaneously. Optionally, an artificial intelligence model can be trained to identify weld defects in the sub-images. When training the AI ​​model, a separate intelligent model can be trained for each type of weld defect, for example, by labeling a large number of images containing that type of weld defect to create a model specifically designed to identify that type. Alternatively, the model can be an intelligent model capable of recognizing multiple weld defects. For example, a model capable of recognizing multiple weld defects can be trained by labeling a large number of images containing various types of weld defects.

[0051] Step S104: For the at least one candidate sub-image, according to the filtering conditions corresponding to the candidate sub-image, the welding defects in the candidate sub-image that meet the filtering conditions are determined as target welding defects.

[0052] Since the detected welding defects may be within the acceptable level of defects, further screening is conducted on the welding defects detected in each candidate sub-image using screening criteria to remove welding defects that are not within the acceptable level.

[0053] In one example, the filtering criteria include at least one of the following: 1) the ratio of the width of the weld defect to the width of the cell it is located in is greater than a first threshold; 2) the ratio of the area of ​​the weld defect to the area of ​​the cell it is located in is greater than a second threshold; 3) the grayscale value of the weld defect is within a preset grayscale range; 4) the confidence level that the weld defect is the target weld defect is greater than a third threshold. The filtering criteria can be any one of the above four conditions, or any combination of the above four conditions. For example, a weld defect is identified as a target weld defect only if it simultaneously satisfies conditions 1) and 4). As another example, a weld defect is identified as a target weld defect only if it simultaneously satisfies conditions 1) and 2).

[0054] In some examples, the identified welding defects may be irregular in shape. Therefore, when determining whether a welding defect meets the above screening condition 1), the width of the weld defect calculated may be the width of the circumscribed rectangle of the identified welding defect, or the diameter of the circumscribed circle, etc., without any restrictions.

[0055] In some examples, when determining whether a welding defect meets the above screening condition 2), the area of ​​the welding defect calculated can be the actual area of ​​the identified welding defect, the area of ​​the circumscribed rectangle of the welding defect, or the area of ​​the circumscribed circle, etc., without any restrictions.

[0056] In this configuration, at least two of the multiple sub-images have different filtering conditions. For example, the at least two sub-images may each correspond to filtering conditions with different content. For instance, one sub-image may use filtering conditions 1) and 4) as described above, while the other sub-image may use filtering conditions 1) and 3) as described above, or conditions 2) and 3) as described above. Alternatively, the at least two sub-images may each correspond to filtering conditions with different thresholds. For instance, both sub-images may use filtering condition 1) as described above, but the values ​​of the first thresholds used in the filtering conditions corresponding to the two sub-images may be set differently.

[0057] After filtering out the target welding defects in the candidate sub-images, since the position of the solar cell in each sub-image within the photovoltaic module is known, the location of each target welding defect can be determined based on that position. The identified target welding defects and their locations are then displayed to the user for quick and easy location.

[0058] In this embodiment of the disclosure, the image with the battery cell array is first segmented to obtain sub-images with at least one battery cell each. Then, welding defect detection is performed on different sub-images. After obtaining candidate sub-images with welding defects, different screening conditions can be conveniently used to screen the welding defects in different candidate sub-images. This satisfies the different requirements for welding defects of battery cells in different regions and groups, making it convenient to match different demand scenarios. This allows for the screening of welding defects in different regions and groups with higher accuracy.

[0059] Optionally, the plurality of sub-images are divided into at least two groups of sub-images, wherein the sub-images in the same group have the same filtering conditions, and the sub-images in different groups have different filtering conditions. Optionally, the plurality of sub-images are divided into at least two groups based on the position of the battery cells in the sub-images within the battery cell array. Optionally, in the at least two groups of sub-images, the sub-images in at least one group are adjacent to each other in the battery cell array, that is, at least some of the battery cells in each group are arranged in a regional pattern. Each group of battery cells may include at least one connected battery cell region.

[0060] In one example, the multiple sets of sub-images are arranged sequentially along the direction parallel to the solder ribbon on the cell array, meaning that cells in the same column are in the same group, while at least two cells in the same row are in different groups. Optionally, a group of cells includes at least two columns of cells. Optionally, the number of columns of cells contained in different groups of cells can be the same or different.

[0061] In one example, the criteria for judging welding defects differ between battery cells located on the cut surface and those located on the non-cut surface. Therefore, among the multiple sub-images, the first sub-image containing battery cells located on the cut surface of the battery cell array and the second sub-image containing battery cells located on the non-cut surface of the battery cell array are grouped into different groups. Optionally, if the filtering conditions corresponding to the first and second sub-images both include condition 1) of the above-mentioned filtering conditions, the first threshold value in the filtering conditions corresponding to the first sub-image can be set to be greater than the first threshold value in the filtering conditions corresponding to the second sub-image. For example, if the first threshold value in the filtering conditions corresponding to the first sub-image is set to 30%, that is, welding defects whose width is not greater than 30% of the width of the battery cell are not considered as target welding defects; and if the first threshold value in the filtering conditions corresponding to the second sub-image is set to 15%, that is, welding defects whose width is not greater than 15% of the width of the battery cell are not considered as target welding defects.

[0062] Similarly, optionally, if the filtering conditions corresponding to the first sub-image and the second sub-image both include condition 2) of the above filtering conditions, the second threshold value in the filtering conditions corresponding to the first sub-image can be set to be greater than the second threshold value in the filtering conditions corresponding to the second sub-image.

[0063] Optionally, the cell array is provided with busbars for electrically connecting the solder strips on at least two cells that are electrically connected to the busbars. The busbars are typically tin-plated copper strips to provide conductivity. In some examples, the requirements for solder defects on cells close to or electrically connected to the busbars differ from those on cells farther from the busbars. Therefore, optionally, in the plurality of sub-images, a third sub-image containing cells in contact with the busbars or at a distance no greater than a preset distance from the busbars, and a fourth sub-image containing cells at a distance greater than the preset distance from the busbars, are grouped into different groups.

[0064] like Figure 3 As shown, Figure 3 This is a schematic diagram of another embodiment of the battery cell assembly in this disclosure. Figure 3In the cell array 30, a busbar is provided in each of the left column region 31, the middle region 32, and the right column region 33, for a total of three busbars. Therefore, the cell array 30 is divided into two groups: one group contains cells located in the left column region 31, the right column region 32, and the middle region 33; the other group contains cells located in region 34 between the left column region 31 and the middle region 32, and cells located in region 35 between the middle region 32 and the right column region 33. Different screening conditions are applied to these two groups of cells.

[0065] In one example, the plurality of sub-images are divided into at least two groups of sub-images based on the chamfer direction of the battery cells within the sub-images. For example, in the plurality of sub-images, all battery cells with a first chamfer direction may be located in the same group, or divided into at least two groups; all battery cells with a second chamfer direction may be located in the same group, or divided into at least two groups. Optionally, battery cells with the first chamfer direction and battery cells with the second chamfer direction are divided into different groups. For example, Figure 2 In the cell array shown, the chamfers of the cells in the first row all face to the left in the diagram, for example, chamfer 241 of cell 24 in the first row. The chamfers of the cells in the second row all face to the right in the diagram, for example, chamfer 231 of cell 23 in the second row.

[0066] In one example, the plurality of sub-images are divided into at least two groups of sub-images based on the chamfer direction of the battery cells in the sub-images and the position of the battery cells in the sub-images within the battery cell array. For example, after dividing the plurality of sub-images into at least two groups based on the position of the battery cells in the sub-images within the battery cell array as described in the above example, each group of sub-images is further divided into two groups based on the chamfer direction.

[0067] This disclosure also provides a device for testing solar cells. For example... Figure 4 As shown, Figure 4 This is a schematic structural block diagram of one embodiment of the solar cell testing apparatus disclosed herein. The solar cell testing apparatus 400 includes:

[0068] The acquisition module 401 is used to acquire a target image, the target image containing a photovoltaic module, the photovoltaic module containing a cell array formed by multiple cells.

[0069] The segmentation module 402 is used to segment the target image to obtain multiple sub-images, wherein each sub-image contains at least one of the battery cells.

[0070] The detection module 403 is used to detect welding defects in the plurality of sub-images respectively, and to obtain at least one candidate sub-image with welding defects.

[0071] The filtering module 404 is used to determine, based on the filtering conditions corresponding to the at least one candidate sub-image, welding defects in the candidate sub-image that meet the filtering conditions as target welding defects; wherein, at least two of the multiple sub-images have different filtering conditions.

[0072] Optionally, the plurality of sub-images are divided into at least two groups of sub-images, wherein the sub-images in the same group of sub-images have the same filtering conditions, and the sub-images in different groups of sub-images have different filtering conditions.

[0073] Optionally, the plurality of sub-images are divided into at least two groups according to the position of the battery cells in the sub-images in the battery cell array.

[0074] Optionally, solder strips are formed on the battery cell array, and the at least two groups are arranged in a direction parallel to the solder strips.

[0075] Optionally, among the plurality of sub-images, a first sub-image having battery cells located at the cut surface of the battery cell array and a second sub-image having battery cells located at the non-cut surface of the battery cell array are located in different groups.

[0076] Optionally, the battery cell array is provided with a busbar for electrically connecting at least two battery cells that are electrically connected to the busbar;

[0077] Among the multiple sub-images, the third sub-image, which has a battery cell that is in contact with the busbar or at a distance from the busbar that is no greater than a preset distance, is in a different group from the fourth sub-image, which has a battery cell that is at a distance from the busbar that is greater than the preset distance.

[0078] Optionally, the plurality of sub-images are divided into at least two groups of sub-images based on the chamfer direction of the battery cells in the sub-images.

[0079] Optionally, in the at least two sets of sub-images, each sub-image in at least one set of sub-images is adjacent to the next in the cell array.

[0080] Optionally, the step of detecting welding defects in the plurality of sub-images includes:

[0081] Welding defects are detected in the multiple sub-images using artificial intelligence algorithms.

[0082] Optionally, the filtering criteria include at least one of the following:

[0083] The ratio of the width of the welding defect to the width of the battery cell it belongs to is greater than a first threshold.

[0084] The ratio of the area of ​​the welding defect to the area of ​​the cell in which it is located is greater than the second threshold.

[0085] The grayscale value of the welding defect is within a preset grayscale range;

[0086] The welding defect is defined as a target welding defect whose confidence level is greater than a third threshold.

[0087] This disclosure also provides a testing device for battery cells, such as... Figure 5 As shown, Figure 5 This is a schematic structural block diagram of one embodiment of the solar cell testing device according to the present disclosure. The solar cell testing device 500 includes at least one memory 501 and at least one processor 502.

[0088] Processor 502 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0089] Memory 501 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage devices. ROM may store static data or instructions required by processor 502 or other modules of the computer. Permanent storage devices may be read-write storage devices. Permanent storage devices may be non-volatile storage devices that retain stored instructions and data even when the computer is powered off. In some embodiments, permanent storage devices use mass storage devices (e.g., magnetic or optical disks, flash memory) as permanent storage devices. In other embodiments, permanent storage devices may be removable storage devices (e.g., floppy disks, optical drives). System memory may be a read-write storage device or a volatile read-write storage device, such as dynamic random access memory. System memory may store some or all of the instructions and data required by the processor during operation. Furthermore, memory 61 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, memory 510 may include a removable storage device that is readable and / or writable, such as a laser disc (CD), a read-only digital multifunction optical disc (e.g., DVD-ROM, dual-layer DVD-ROM), a read-only Blu-ray disc, an ultra-high density optical disc, a flash memory card (e.g., SD card, mini SD card, Micro-SD card, etc.), a magnetic floppy disk, etc. Computer-readable storage media do not contain carrier waves or transient electronic signals transmitted wirelessly or via wired connections.

[0090] The memory 501 stores executable code, which, when processed by the processor 502, can cause the processor 502 to execute part or all of the methods described above.

[0091] In one example, the solar cell detection device includes an AI server and a data server. The AI ​​server includes a first memory and a first processor, while the data server includes a second memory and a second processor. The first memory in the AI ​​server stores executable code. When the executable code is processed by the first processor, it can execute steps S101 to S103 and send the candidate sub-images obtained in step S103 to the data server. The second memory in the data server stores executable code. When the executable code is processed by the second processor, it can execute step S104 based on the received candidate sub-images.

[0092] Furthermore, the method according to this disclosure can also be implemented as a computer program or computer program product, which includes computer program code instructions for performing some or all of the steps in the method described above.

[0093] Alternatively, this disclosure may be implemented as a computer-readable storage medium (or a non-transitory machine-readable storage medium or a machine-readable storage medium) storing executable code (or computer program or computer instruction code) that, when executed by one or more processors of an electronic device (or server, etc.), causes the processor to perform some or all of the steps of the methods described above according to this disclosure.

[0094] The above descriptions are merely preferred embodiments of this disclosure. The descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical applications, or improvements to the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein. Various elements in the embodiments or examples may be omitted or replaced by their equivalents. Furthermore, the steps may be performed in a different order than that described in this disclosure. Further, various elements in the embodiments or examples may be combined in various ways. Importantly, as technology evolves, many elements described herein can be replaced by equivalents that appear after this disclosure. That is, any equivalent structural or procedural transformations made based on the claims, description, and drawings of this disclosure, or direct or indirect applications in other related technical fields, are similarly included within the scope of protection of the claims of this patent.

Claims

1. A method for testing battery cells, characterized in that, include: Acquire a target image, wherein the target image contains a photovoltaic module, and the photovoltaic module comprises a cell array formed by multiple cells; The target image is segmented to obtain multiple sub-images, wherein each sub-image contains at least one of the battery cells; Welding defect detection is performed on the multiple sub-images respectively to obtain at least one candidate sub-image with welding defects; For the at least one candidate sub-image, according to the filtering conditions corresponding to the candidate sub-image, welding defects in the candidate sub-image that meet the filtering conditions are identified as target welding defects; wherein, at least two sub-images in the plurality of sub-images correspond to different filtering conditions; the plurality of sub-images are divided into at least two groups of sub-images, wherein the sub-images in the same group of sub-images correspond to the same filtering conditions, and the sub-images in different groups of sub-images correspond to different filtering conditions; in the at least two groups of sub-images, each sub-image in at least one group of sub-images is adjacent to each other in the solar cell array.

2. The method according to claim 1, characterized in that, The plurality of sub-images are divided into at least two groups according to the position of the battery cells in the sub-images within the battery cell array.

3. The method according to claim 2, characterized in that, The battery cell array has solder strips formed on it, and the at least two groups are arranged in a direction parallel to the solder strips.

4. The method according to claim 2, characterized in that, Among the plurality of sub-images, a first sub-image having battery cells located at the cut surface of the battery cell array and a second sub-image having battery cells located at the non-cut surface of the battery cell array are located in different groups.

5. The method according to claim 2, characterized in that, The battery cell array is provided with a bus bar for electrically connecting at least two battery cells that are electrically connected to the bus bar; Among the multiple sub-images, the third sub-image, which has a battery cell that is in contact with the busbar or at a distance from the busbar that is no greater than a preset distance, is in a different group from the fourth sub-image, which has a battery cell that is at a distance from the busbar that is greater than the preset distance.

6. The method according to claim 1, characterized in that, The plurality of sub-images are divided into at least two groups of sub-images based on the chamfering direction of the battery cells in the sub-images.

7. The method according to any one of claims 1 to 6, characterized in that, The step of detecting welding defects in the plurality of sub-images includes: Welding defects are detected in the multiple sub-images using artificial intelligence algorithms.

8. The method according to any one of claims 1 to 6, characterized in that, The filtering criteria include at least one of the following: The ratio of the width of the welding defect to the width of the battery cell it belongs to is greater than a first threshold. The ratio of the area of ​​the welding defect to the area of ​​the cell in which it is located is greater than the second threshold. The grayscale value of the welding defect is within a preset grayscale range; The welding defect is defined as a target welding defect whose confidence level is greater than a third threshold.

9. A device for detecting battery cells, characterized in that, include: An acquisition module is used to acquire a target image, wherein the target image contains a photovoltaic module, and the photovoltaic module contains a cell array formed by multiple cells. A segmentation module is used to segment the target image to obtain multiple sub-images, wherein each sub-image contains at least one of the battery cells; The detection module is used to detect welding defects in the multiple sub-images respectively, and obtain at least one candidate sub-image with welding defects; A filtering module is used to identify welding defects in at least one candidate sub-image as target welding defects based on the filtering conditions corresponding to the candidate sub-image; wherein, at least two sub-images in the plurality of sub-images have different filtering conditions; the plurality of sub-images are divided into at least two groups of sub-images, wherein the sub-images in the same group of sub-images have the same filtering conditions, and the sub-images in different groups of sub-images have different filtering conditions; in the at least two groups of sub-images, each sub-image in at least one group of sub-images is adjacent to the next to the next in the solar cell array.

10. A testing device for solar cells, characterized in that, It includes a memory and a processor, wherein the memory stores executable code, and when the executable code is processed by the processor, the processor can perform the method of any one of claims 1 to 8.

11. A computer-readable storage medium having executable code stored thereon, which, when executed by a processor, causes the processor to perform the method as claimed in any one of claims 1 to 8.

Images