A method and apparatus for detecting hot melt area of a mylar process of a lithium battery pack

Abstract

This invention relates to the field of lithium battery manufacturing technology, specifically to a method and apparatus for detecting the hot melt area in the Mylar process of a lithium battery pack. The method includes acquiring an original image; performing region localization based on the original image to obtain a region image; performing multiple hot melt area extractions on the region image based on a deep learning model: each time hot melt area extraction is performed on the region image, each adhesive dot is extracted to a corresponding hot melt area, and each hot melt area has a corresponding confidence level; after multiple hot melt area extractions on the region image, the hot melt area with the highest confidence level is selected for each adhesive dot as the final extracted hot melt area for that dot; the hot melt area is binarized to obtain a binarized region, and connected component extraction is performed on the binarized region. This method can effectively filter out accidental misjudgments, ensuring that the final extracted hot melt area is stable and reliable, greatly reducing the possibility of good products being misjudged as NG or defective products being passed over.

Description

Technical Field

[0001] This invention relates to the field of lithium battery production technology, specifically to a method and apparatus for detecting the hot melt area of ​​the Mylar process in lithium battery packs. Background Technology

[0002] In the manufacturing process of aluminum-cased lithium batteries, before the bare cell is installed into the aluminum casing, it needs to be wrapped with a layer of polyester polymer insulating film (Mylar film) to achieve insulation protection. During the wrapping process, in order to prevent the Mylar film from loosening or falling off, the equipment uses heating strips (heat sealing blades) to heat and pressurize the Mylar film at specific locations (usually around the bottom of the battery or at the top edge), so that it melts and adheres to the base plate or casing.

[0003] To ensure the fixation strength of the Mylar film, prevent welding "explosions" and "film jamming" in subsequent processes, and guarantee compliance with insulation distance requirements, it is necessary to perform hot melt area inspection. Hot melt area inspection refers to the process of measuring and determining the size of the effective weld area after the Mylar film is fixed to the battery casing or base plate by hot melt welding, using machine vision (CCD) or other inspection methods in the lithium battery pack Mylar film process.

[0004] Current technology uses grayscale control to detect melting point area, but this method has poor stability in its results. A good product typically has a valid melting point area, but there's a significant chance it might be detected as having an invalid melting point area. When the detected melting point area is invalid, the good product is mistakenly rejected. Similarly, a defective product typically has an invalid melting point area, but there's a significant chance it might be detected as having a valid melting point area. When the detected melting point area is valid, the defective product is allowed to pass through. Summary of the Invention

[0005] The purpose of this invention is to provide a method and apparatus for detecting the hot melt area in the Mylar process of lithium battery packs. By extracting the hot melt area multiple times, multiple hot melt area extraction results and their corresponding confidence levels are obtained. The hot melt area extraction result with the highest confidence level is selected, which can effectively filter out accidental misjudgments, so that the finally extracted hot melt area is stable and reliable, greatly reducing the situation where good products are misjudged as NG (over-detection) or defective products are let go (missed detection).

[0006] In a first aspect of the present invention, a method for detecting the hot melt area of ​​a Mylar process in a lithium battery pack is provided, comprising:

[0007] S1. Obtain the original image, and perform region localization based on the original image to obtain the region image;

[0008] S2. Multiple hot melt region extractions are performed on the region image based on a deep learning model: Each time the hot melt region is extracted from the region image, each glue dot is extracted into a corresponding hot melt region, and each hot melt region has a corresponding confidence level.

[0009] After extracting the hot melt region multiple times from the region image, the hot melt region with the highest confidence is selected for each glue point as the final extracted hot melt region for the corresponding glue point.

[0010] S3. Perform binarization on the hot-melt region to obtain a binarized region, and then obtain the connected components from the binarized region;

[0011] S4. Obtain the actual area of ​​the connected component and determine whether the actual area of ​​the connected component is greater than the first area threshold or less than the second area threshold. If the actual area of ​​the connected component is greater than the first area threshold, the hot-melt area is determined to be too large; if the actual area of ​​the connected component is less than the second area threshold, the hot-melt area is determined to be too small; otherwise, the hot-melt area is determined to be qualified.

[0012] As a preferred embodiment of the present invention, in S1, obtaining a region image by performing region localization based on the original image specifically includes:

[0013] S11. Detect the outer edge of the lithium battery top cover, and obtain the intersection point of the outer edge of the top cover and the fixed longitudinal line based on the lithium battery fitting to serve as the coarse positioning point.

[0014] S12. Grasp the side edge of the lithium battery top cover based on coarse positioning points;

[0015] S13. Obtain the intersection of the outer edge of the top cover and the side edge of the top cover as the fine positioning point;

[0016] S14. Based on the precise positioning points, obtain the region of interest surrounding the edge of the lithium battery, and crop the region of interest from the original image to obtain the region image.

[0017] In a preferred embodiment of the present invention, in S2, the region image is extracted three times based on a deep learning model to extract the thermal melting region.

[0018] As a preferred embodiment of the present invention, in S3, a mask method is used to simultaneously binarize several hot-melt regions on the region image, and a mask method is used to simultaneously obtain connected components from several binarized regions on the region image.

[0019] As a preferred embodiment of the present invention, binarizing the hot-melt region to obtain a binarized region specifically includes:

[0020] The gray values ​​of pixels in the hot-melt region that have a gray value greater than or equal to the gray value threshold are set as the first gray value, and the gray values ​​of pixels in the hot-melt region that have a gray value less than the gray value threshold are set as the second gray value.

[0021] As a preferred embodiment of the present invention, in S4, obtaining the actual area of ​​the connected region specifically includes:

[0022] Get the pixel area of ​​the connected component and convert the pixel area to the actual area.

[0023] As a preferred embodiment of the present invention, if the hot-melt region cannot be extracted in S2 or the connected region cannot be obtained in S3, the hot-melt area is 0.

[0024] In a second aspect of the present invention, a device for detecting the hot melt area of ​​a lithium battery pack Mylar process is provided, comprising:

[0025] The region image acquisition module is configured to acquire the original image and perform region localization based on the original image to obtain the region image.

[0026] The hot melt region extraction module is configured to perform multiple hot melt region extractions on the region image based on a deep learning model: each time the hot melt region is extracted from the region image, each glue dot is extracted to a corresponding hot melt region, and each hot melt region has a corresponding confidence level.

[0027] After extracting the hot melt region multiple times from the region image, the hot melt region with the highest confidence is selected for each glue point as the final extracted hot melt region for the corresponding glue point.

[0028] The connected component acquisition module is configured to perform binarization processing on the hot-melt region to obtain a binary region, and then acquire the connected components from the binary region.

[0029] The hot melt area determination module is configured to obtain the actual area of ​​the connected region and determine whether the actual area of ​​the connected region is greater than a first area threshold or less than a second area threshold. When the actual area of ​​the connected region is greater than the first area threshold, the hot melt area is determined to be too large; when the actual area of ​​the connected region is less than the second area threshold, the hot melt area is determined to be too small; otherwise, the hot melt area is determined to be qualified.

[0030] In a third aspect of the present invention, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to the first aspect.

[0031] In a fourth aspect of the present invention, an electronic device is provided, including one or more processors and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the method provided in the first aspect.

[0032] In summary, the present invention has the following beneficial effects:

[0033] The hot melt area detection method and apparatus of this invention obtain multiple hot melt area extraction results and their corresponding confidence levels through multiple hot melt area extractions. By selecting the hot melt area extraction result with the highest confidence level, it can effectively filter out accidental misjudgments, making the finally extracted hot melt area stable and reliable. This greatly reduces the situation where good products are misjudged as NG (over-detection) or defective products are let go (missed detection).

[0034] Further or more detailed beneficial effects will be described in conjunction with specific embodiments in the detailed implementation. Attached Figure Description

[0035] Figure 1 A flowchart of the hot melt area detection method for the Mylar process of lithium battery pack according to an embodiment of the present invention is shown;

[0036] Figure 2 A block diagram of a hot melt area detection device for the Mylar process of a lithium battery pack according to an embodiment of the present invention is shown.

[0037] Figure 3 A block diagram of an electronic device according to an embodiment of the present invention is shown. Detailed Implementation

[0038] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the invention. It should be understood that the accompanying drawings and embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the invention.

[0039] In the description of embodiments of the present invention, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0040] Figure 1A flowchart illustrating a method for detecting the hot melt area in the Mylar process of a lithium battery pack according to an embodiment of the present invention is shown. The method includes:

[0041] S1. Obtain the original image, and perform region localization based on the original image to obtain the region image.

[0042] This embodiment acquires raw images using an image acquisition module. The image acquisition module is used to acquire surface images of a lithium battery cell coated with a Mylar film. The module includes a first imaging component arranged in a first direction (e.g., the long side direction of the lithium battery cell) and a second imaging component arranged in a second direction (e.g., the short side direction of the lithium battery cell), and also includes illumination components corresponding to the first and second imaging components. For the same lithium battery, the image acquisition module can acquire four raw images simultaneously, each corresponding to one side of the lithium battery.

[0043] Taking an original image as an example, in this embodiment, obtaining a region image by locating a region based on the original image specifically includes:

[0044] S11. Detect the outer edge of the lithium battery top cover, and obtain the intersection point of the outer edge of the top cover and the fixed longitudinal line based on the lithium battery fitting to serve as the coarse positioning point.

[0045] S12. Grasp the side edge of the lithium battery top cover based on coarse positioning points;

[0046] S13. Obtain the intersection of the outer edge of the top cover and the side edge of the top cover as the fine positioning point;

[0047] S14. Based on the precise positioning points, obtain the region of interest surrounding the edge of the lithium battery, and crop the region of interest from the original image to obtain the region image.

[0048] In this embodiment, a region image is obtained through an image processing module. After obtaining the original image, the image processing module processes the original image through steps S11 to S14 to obtain the region image. The original image may include other things such as a conveyor belt and background support in addition to the lithium battery, while the region image only includes the lithium battery, which facilitates the accurate extraction of the hot-melt area in the subsequent process.

[0049] In this embodiment, the image processing module also performs grayscale processing on the regional image.

[0050] S2. Multiple hot melt region extractions are performed on the region image based on a deep learning model: Each time the hot melt region is extracted from the region image, each glue dot is extracted into a corresponding hot melt region, and each hot melt region has a corresponding confidence level.

[0051] After extracting the hot melt region multiple times from the regional image, the hot melt region with the highest confidence is selected for each glue dot as the final extracted hot melt region for that glue dot.

[0052] In this embodiment, step S2 is executed by the image processing module.

[0053] Taking a region image as an example, in this embodiment, the region image is extracted three times based on a deep learning model. Assume there are two glue dots in the region image: glue dot A and glue dot B.

[0054] After the deep learning model extracts the hot melt region from the image for the first time, it will obtain the hot melt region A-1 (for glue point A) and confidence level a-1, as well as the hot melt region B-1 (for glue point B) and confidence level b-1.

[0055] After the deep learning model extracts the hot melt region from the image for the second time, it will obtain the hot melt region A-2 (for glue point A) and confidence level a-2, as well as the hot melt region B-2 (for glue point B) and confidence level b-2.

[0056] After the deep learning model extracts the hot melt region for the third time from the image, it will obtain hot melt region A-3 (for glue point A) and confidence level a-3, as well as hot melt region B-3 (for glue point B) and confidence level b-3.

[0057] In this embodiment, the hot-melt region can be represented by two pixel coordinates, such as D1(x1, y1) and D2(x2, y2), representing the region on the image where the X coordinate of a pixel is from x1 to x2 and the Y coordinate of a pixel is from y1 to y2. When the deep learning model obtains a hot-melt region, it automatically obtains the confidence level of the hot-melt region. The confidence level can be understood as the probability that the obtained hot-melt region is indeed a hot-melt region.

[0058] Assuming that confidence level a-2 is greater than confidence level a-1, which is greater than confidence level a-3, then the hot-melt region A-2 is taken as the final hot-melt region extracted from adhesive point A. Further assuming that confidence level b-3 is greater than confidence level b-2, which is greater than confidence level b-1, then the hot-melt region B-3 is taken as the final hot-melt region extracted from adhesive point B.

[0059] If a region image cannot capture the hot melt area, that is, if no hot melt area is captured, then the hot melt area is directly determined to be 0 (i.e., no glue dots).

[0060] S3. Binarize the hot-melt region to obtain a binarized region, and then obtain the connected components from the binarized region.

[0061] In this embodiment, step S3 is executed by the image processing module.

[0062] In this embodiment, a masking method is used to simultaneously binarize several hot-melt regions on the region image, and then a masking method is used to simultaneously obtain connected components from these binarized regions. The positions of the binarized regions on the region image are exactly the same as the positions of the hot-melt regions. First, this method eliminates the need to crop the hot-melt regions from the region image; second, it allows all hot-melt regions to be binarized simultaneously, and all binarized regions to have their connected components obtained simultaneously, effectively improving image processing efficiency.

[0063] Taking a region image as an example, assuming that the hot melt regions obtained in step S2 are hot melt region A-2 and hot melt region B-3, this step will first perform binarization processing on hot melt region A-2 and hot melt region B-3 simultaneously through masking.

[0064] In this embodiment, binarizing the hot-melt region to obtain a binarized region specifically includes:

[0065] The gray values ​​of pixels in the hot-melt region that have a gray value greater than or equal to the gray value threshold are set as the first gray value, and the gray values ​​of pixels in the hot-melt region that have a gray value less than the gray value threshold are set as the second gray value.

[0066] The Mylar film background is relatively bright (grayscale value approximately 200), while the hot melt adhesive dots appear dark (grayscale value approximately 50-80) due to the material. This embodiment assumes a grayscale threshold of 150, a first grayscale value of 0, and a second grayscale value of 255. Taking a pixel within the hot melt area as an example, when the grayscale value of that pixel is greater than or equal to 150, the grayscale value of that pixel is set to 0; when the grayscale value of that pixel is less than 150, the grayscale value of that pixel is set to 255. The other pixels in the hot melt area are processed similarly, ultimately resulting in a binarized region corresponding to that hot melt area. Binarizing hot melt area A-2 yields binarized region A, and binarizing hot melt area B-3 yields binarized region B.

[0067] After the hot-melt region is binarized, the image processing module performs 8-connected component analysis on the binarized region to obtain the connected components. Typically, binarized region A will yield connected component A, and binarized region B will yield connected component B.

[0068] If no connected components can be obtained from the entire binarized region, then the hot melt area is determined to be 0 (i.e., there are no glue dots).

[0069] S4. Obtain the actual area of ​​the connected component and determine whether the actual area of ​​the connected component is greater than the first area threshold or less than the second area threshold. If the actual area of ​​the connected component is greater than the first area threshold, the hot-melt area is determined to be too large; if the actual area of ​​the connected component is less than the second area threshold, the hot-melt area is determined to be too small; otherwise, the hot-melt area is determined to be qualified.

[0070] Taking a connected region A as an example, in this embodiment, obtaining the actual area of ​​the connected region specifically includes: obtaining the pixel area of ​​the connected region and converting the pixel area into the actual area. Assuming that connected region A includes 400 pixels, each pixel being 0.0025 square millimeters, then the actual area of ​​connected region A is 1 square millimeter.

[0071] Assume the first area threshold is 1.2 square millimeters and the second area threshold is 0.8 square millimeters. If the actual area of ​​connected region A is greater than 1.2 square millimeters, the hot-melt area is determined to be too large; if the actual area of ​​connected region A is less than 0.8 square millimeters, the hot-melt area is determined to be too small; otherwise, the hot-melt area is determined to be acceptable. In this embodiment, if the actual area of ​​connected region A is 1 square millimeter, the hot-melt area is determined to be acceptable.

[0072] For a given region image, if some connected components determine the molten area to be acceptable while others determine it to be too large, then the final determination for that region image is that the molten area is too large. Similarly, for a lithium battery, if some regions in an image are determined to have an excessively large molten area while others are deemed to have an acceptable molten area, then the final determination for that lithium battery is that its molten area is too large.

[0073] Figure 2 A block diagram of a hot melt area detection device for the Mylar process of a lithium battery pack according to an embodiment of the present invention is shown. The device includes:

[0074] The region image acquisition module is configured to acquire the original image and perform region localization based on the original image to obtain the region image.

[0075] The hot melt region extraction module is configured to perform multiple hot melt region extractions on the region image based on a deep learning model: each time the hot melt region is extracted from the region image, each glue dot is extracted to a corresponding hot melt region, and each hot melt region has a corresponding confidence level.

[0076] After extracting the hot melt region multiple times from the region image, the hot melt region with the highest confidence is selected for each glue point as the final extracted hot melt region for the corresponding glue point.

[0077] The connected component acquisition module is configured to perform binarization processing on the hot-melt region to obtain a binary region, and then acquire the connected components from the binary region.

[0078] The hot melt area determination module is configured to obtain the actual area of ​​the connected region and determine whether the actual area of ​​the connected region is greater than a first area threshold or less than a second area threshold. When the actual area of ​​the connected region is greater than the first area threshold, the hot melt area is determined to be too large; when the actual area of ​​the connected region is less than the second area threshold, the hot melt area is determined to be too small; otherwise, the hot melt area is determined to be qualified.

[0079] Figure 3 A block diagram of an electronic device 300 according to some embodiments of the present invention is shown. The device 300 includes a processor 301, which performs various appropriate actions and processes based on computer program instructions loaded into random access memory (RAM) 303 according to computer program instructions stored in read-only memory (ROM) 302. Various programs and data required for the operation of the device 300 may also be stored in RAM 303. The processor 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0080] The various processes and procedures described above, such as method 100, can be executed by processor 301. For example, in some embodiments, method 100 may be implemented as a software program tangibly contained in a machine-readable medium. In some embodiments, part or all of the software program may be loaded and / or installed on device 300 via ROM 302. When the software program is loaded into RAM 303 and executed by processor 301, one or more actions of method 100 described above may be performed.

[0081] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload programmable logic devices (CPLDs), and so on.

[0082] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0083] This invention can be a method, apparatus, system, and / or program product. The program product may include a machine-readable storage medium on which machine-readable program instructions for performing various aspects of the invention are loaded. The machine-readable program instructions described herein can be downloaded from the machine-readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the machine-readable program instructions from the network and forwards them for storage in the machine-readable storage medium of the respective computing / processing device.

[0084] Machine program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The machine-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the machine-readable program instructions. This electronic circuitry can execute the machine-readable program instructions to implement various aspects of the invention.

[0085] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, although the operations are depicted in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the invention. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0086] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method for detecting the hot melt area in the Mylar process of a lithium battery pack, characterized in that, include: S1. Obtain the original image, and perform region localization based on the original image to obtain a region image; S2. Multiple hot melt region extractions are performed on the region image based on a deep learning model: Each time the hot melt region is extracted from the region image, each glue dot is extracted into a corresponding hot melt region, and each hot melt region has a corresponding confidence level. After extracting the hot melt region multiple times from the region image, the hot melt region with the highest confidence is selected for each glue point as the final extracted hot melt region for the corresponding glue point. S3. Perform binarization processing on the hot-melt region to obtain a binarized region, and obtain the connected components of the binarized region; S4. Obtain the actual area of ​​the connected region, and determine whether the actual area of ​​the connected region is greater than a first area threshold or less than a second area threshold. When the actual area of ​​the connected region is greater than the first area threshold, it is determined that the hot-melt area is too large; when the actual area of ​​the connected region is less than the second area threshold, it is determined that the hot-melt area is too small; otherwise, it is determined that the hot-melt area is qualified.

2. The method according to claim 1, characterized in that, In step S1, obtaining a region image by locating a region based on the original image specifically includes: S11. Detect the outer edge of the lithium battery top cover, and obtain the intersection point of the outer edge of the top cover and the fixed longitudinal line based on the lithium battery fitting to serve as the coarse positioning point. S12. Grasp the side edge of the lithium battery top cover based on coarse positioning points; S13. Obtain the intersection of the outer edge of the top cover and the side edge of the top cover as the fine positioning point; S14. Obtain the region of interest surrounding the edge of the lithium battery based on the precise positioning points, and crop the region of interest from the original image to obtain a region image.

3. The method according to claim 1, characterized in that, In step S2, the region image is extracted three times based on a deep learning model to extract the thermal melting region.

4. The method according to claim 1, characterized in that, In step S3, a mask is used to simultaneously binarize several hot-melt regions on the region image, and a mask is used to simultaneously obtain connected components from several binarized regions on the region image.

5. The method according to claim 4, characterized in that, The binarization process for the hot-melt region to obtain the binarized region specifically includes: The gray values ​​of pixels in the hot-melt region that have a gray value greater than or equal to the gray value threshold are set as the first gray value, and the gray values ​​of pixels in the hot-melt region that have a gray value less than the gray value threshold are set as the second gray value.

6. The method according to claim 1, characterized in that, In step S4, obtaining the actual area of ​​the connected component specifically includes: Obtain the pixel area of ​​the connected component and convert the pixel area into the actual area.

7. The method according to claim 1, characterized in that, If no hot-melt region can be extracted in S2 or no connected region can be obtained in S3, then the hot-melt area is 0.

8. A method and apparatus for detecting the hot melt area in the Mylar process of a lithium battery pack, characterized in that, include: The region image acquisition module is configured to acquire an original image and perform region localization based on the original image to obtain a region image. The hot melt region extraction module is configured to perform multiple hot melt region extractions on the region image based on a deep learning model: each time the hot melt region is extracted from the region image, each glue dot is extracted to a corresponding hot melt region, and each hot melt region has a corresponding confidence level. After extracting the hot melt region multiple times from the region image, the hot melt region with the highest confidence is selected for each glue point as the final extracted hot melt region for the corresponding glue point. The connected component acquisition module is configured to perform binarization processing on the hot-melt region to obtain a binarized region, and then acquire the connected component of the binarized region. The hot melt area determination module is configured to obtain the actual area of ​​the connected region and determine whether the actual area of ​​the connected region is greater than a first area threshold or less than a second area threshold. When the actual area of ​​the connected region is greater than the first area threshold, the hot melt area is determined to be too large; when the actual area of ​​the connected region is less than the second area threshold, the hot melt area is determined to be too small; otherwise, the hot melt area is determined to be qualified.

9. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1-7.

10. An electronic device, characterized in that, It includes one or more processors and memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method according to any one of claims 1-7.

Images