WO2024016266A1 - 检测电芯组件的极耳外观的方法与装置、电子设备 - Google Patents

检测电芯组件的极耳外观的方法与装置、电子设备 Download PDF

Info

Publication number
WO2024016266A1
WO2024016266A1 PCT/CN2022/107073 CN2022107073W WO2024016266A1 WO 2024016266 A1 WO2024016266 A1 WO 2024016266A1 CN 2022107073 W CN2022107073 W CN 2022107073W WO 2024016266 A1 WO2024016266 A1 WO 2024016266A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
area
tab
main body
detection
Prior art date
Application number
PCT/CN2022/107073
Other languages
English (en)
French (fr)
Inventor
李璐
王智玉
江冠南
Original Assignee
宁德时代新能源科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 宁德时代新能源科技股份有限公司 filed Critical 宁德时代新能源科技股份有限公司
Priority to PCT/CN2022/107073 priority Critical patent/WO2024016266A1/zh
Priority to EP22888600.8A priority patent/EP4339887A1/en
Priority to CN202280006741.3A priority patent/CN117751381A/zh
Priority to US18/138,501 priority patent/US11915410B2/en
Publication of WO2024016266A1 publication Critical patent/WO2024016266A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G06T7/0008—Industrial image inspection checking presence/absence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G06T7/001—Industrial image inspection using an image reference approach
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00—Geometric image transformations in the plane of the image
    • G06T3/60—Rotation of whole images or parts thereof
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/20—Image enhancement or restoration using local operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/761—Proximity, similarity or dissimilarity measures
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/04—Construction or manufacture in general
    • H01M10/0431—Cells with wound or folded electrodes
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
    • H01M10/4285—Testing apparatus
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M50/00—Constructional details or processes of manufacture of the non-active parts of electrochemical cells other than fuel cells, e.g. hybrid cells
    • H01M50/10—Primary casings; Jackets or wrappings
    • H01M50/102—Primary casings; Jackets or wrappings characterised by their shape or physical structure
    • H01M50/103—Primary casings; Jackets or wrappings characterised by their shape or physical structure prismatic or rectangular
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20112—Image segmentation details
    • G06T2207/20164—Salient point detection; Corner detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20172—Image enhancement details
    • G06T2207/20192—Edge enhancement; Edge preservation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30164—Workpiece; Machine component

Definitions

  • the present disclosure relates to the technical field of battery production, and in particular, to a method and device for detecting the appearance of tabs of a battery core assembly, electronic equipment, computer-readable storage media, and computer program products.
  • rechargeable batteries referring to batteries that can be activated by charging to activate active materials and continue to be used after the battery is discharged, also known as secondary batteries, hereinafter referred to as batteries
  • batteries include a battery box and a series connected battery located in the battery box. and/or multiple battery cells combined in parallel.
  • the battery cell is the smallest unit that provides an energy source in the battery, and the battery cell component is the key component for electrochemical reactions in the battery cell. Defect detection of the tab appearance of the battery cell component is an important part of battery production. .
  • one object of the present disclosure is to provide a method and device, electronic equipment, computer-readable storage media and computer program products for detecting the appearance of tabs of a battery core assembly.
  • An embodiment of the first aspect of the present disclosure provides a method for detecting the appearance of tabs of a battery cell assembly, including: acquiring a detection image including a background area and a battery cell assembly image area, where the battery cell assembly image area includes a main body area and a battery cell assembly image area.
  • each tab stacking area is adjacent to the top or bottom edge of the main body area; determine each root corner point of the multiple tab stacking areas in the detection image; determine the main body area in the detection image Two sides; based on the two sides of the main body area in the detection image, at least one reference edge line is determined in the detection image; and, based on each root corner point of the plurality of tab stacking areas in the detection image, and at least A reference edge line to determine the detection result information of the tab appearance.
  • the method of the embodiment of the present disclosure uses computer vision technology to detect the appearance of the stacked tabs of the battery core assembly based on the image of the battery core assembly. Compared with the manual measurement or detection methods with the help of special fixtures in the related art, it not only The detection efficiency is high, applicable to a wide range of product specifications, and can significantly improve the accuracy of detection, thereby improving the production yield of cell components and batteries.
  • obtaining the detection image includes: performing binarization processing on the detection image to obtain a binary image, in which the battery cell component image area serves as the area of interest of the binary image.
  • the detection image is binarized to facilitate accurate extraction of the battery component image area in subsequent steps.
  • determining each root corner point of the plurality of tab stacking areas in the detection image includes: in the binary image, based on the template image and the local search area correspondingly configured for each root corner point, in Each local search area performs matching and positioning on the corresponding template image; and, based on the matching and positioning results of the template image in each local search area, determines each root corner point of the multiple tab stacking areas in the binary image.
  • the template image is obtained based on the cell component image template.
  • the local search area is a small-scale search area preset for the binary image. Matching and positioning the template image in the local search area can greatly reduce the amount of matching calculations and improve the speed and accuracy of matching calculations.
  • determining the two sides of the main body area in the detection image includes: extracting the battery cell component image area from the binary image; rolling the battery cell component image area along the row direction in the binary image by one pixel column to obtain a comparison image; based on the grayscale difference between the binary image and the comparison image, determine multiple edge pixels in the battery cell component image area in the binary image; and filter out the pixels in the cell component image area from the multiple edge pixels.
  • the edge pixels arranged in one direction but with a distance smaller than the distance threshold are used to obtain two sides of the main body area in the binary image, where the first direction is approximately the same as the extension direction of the top and bottom edges of the main body area.
  • edge pixels that are irrelevant to the two sides of the main body area can be filtered out, and on the other hand, the loss of image features or errors caused by poor hardware or shooting environment can be avoided as much as possible, thereby further improving Detection accuracy.
  • determining at least one reference edge line in the detection image based on two sides of the main body area in the detection image includes: performing straight line fitting on the two sides of the main body area in the binary image. , obtain two fitting sides extending straight lines; and, based on the two fitting sides, determine at least one reference edge line in the binary image. Determining the reference edge line based on the fitted side obtained by straight line fitting can further improve the accuracy of detection.
  • extracting the battery cell component image area from the binary image includes: extracting the battery cell component from the binary image based on at least one of a threshold segmentation algorithm, an image segmentation algorithm, and a maximum connected domain algorithm. image area.
  • obtaining the detection image includes: performing rotation correction on the detection image so that the top and bottom edges of the main body area are generally parallel to the row direction, and the two side edges of the main body area are generally orthogonal to the row direction. .
  • the exposure time of the detection image is no less than 8000 microseconds and no more than 12000 microseconds. Obtaining a high-exposure detection image makes the grayscale of the battery cell component image area and the background area significantly different, which is helpful to further improve the accuracy of detection.
  • the detection result information of the tab appearance includes: the distance between two root corner points of each tab stacking area; and/or the pass or fail evaluation result of the tab appearance.
  • the cell assembly image area includes two tab stacking areas, namely a positive tab stacking area and a negative tab stacking area, wherein the positive tab stacking area and the negative tab stacking area are common with the main body area.
  • the top edge or the bottom edge of the positive electrode tab stacking area is adjacent to the top edge of the main body area, and the negative electrode tab stacking area is adjacent to the bottom edge of the main body area.
  • An embodiment of the second aspect of the present disclosure provides a device for detecting the appearance of tabs of a battery cell assembly, including: an acquisition unit configured to acquire a detection image including a background area and a battery cell assembly image area, wherein the battery cell assembly image The area includes a main body area and a plurality of tab stacking areas, each tab stacking area being adjacent to the top or bottom edge of the main body area; the first determining unit is configured to determine each of the plurality of tab stacking areas in the detection image. root corner points; the second determination unit is configured to determine two sides of the main body area in the detection image; the third determination unit is configured to determine at least two sides of the main body area in the detection image based on the two sides of the main body area in the detection image. a reference edge line; and a fourth determination unit configured to determine the detection result information of the tab appearance based on each root corner point of the plurality of tab stacking areas in the detection image and at least one reference edge line.
  • An embodiment of the third aspect of the present disclosure provides a computer-readable storage medium storing computer instructions configured to cause a computer to perform the method of the foregoing aspect.
  • An embodiment of the third aspect of the present disclosure provides a computer program product, including a computer program.
  • the computer program implements the method of the foregoing aspect when executed by a processor.
  • Figure 1A is a simplified schematic diagram of the front structure of a wound battery cell assembly
  • Figure 1B is a simplified schematic diagram of the top structure of a wound battery cell assembly
  • Figure 2 is a schematic flowchart of a method for detecting the appearance of tabs of a battery core assembly according to some embodiments of the present disclosure
  • Figure 3 is a schematic flowchart of a method for detecting the appearance of tabs of a battery core assembly according to some embodiments of the present disclosure
  • Figure 4 is a schematic flowchart of a method for detecting the appearance of tabs of a battery core assembly according to some embodiments of the present disclosure
  • FIG. 5 is a structural block diagram of a device for detecting the appearance of tabs of a battery core assembly according to some embodiments of the present disclosure.
  • an embodiment means that a particular feature, structure or characteristic described in connection with the embodiment can be included in at least one embodiment of the present disclosure.
  • the appearances of this phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those skilled in the art understand, both explicitly and implicitly, that the embodiments described herein may be combined with other embodiments.
  • multiple refers to more than two (including two).
  • multiple groups refers to two or more groups (including two groups), and “multiple pieces” refers to It is more than two pieces (including two pieces).
  • the laminated battery cell assembly includes a plurality of positive electrode pieces and a plurality of negative electrode pieces alternately stacked, and a separator disposed between any adjacent positive electrode pieces and negative electrode pieces.
  • the wound battery cell assembly includes a positive electrode piece, a negative electrode piece, and a separator located between the positive electrode piece and the negative electrode piece that are rolled after being laminated.
  • the cell components are soaked by the electrolyte, and the lithium ions use the electrolyte as a medium to move between the negative electrode and the positive electrode, so that the battery cell can be charged and charged.
  • the function of the separator is to allow lithium ions to pass freely without allowing electrons to pass, thus preventing a short circuit between the negative and positive electrodes of the battery cell through the electrolyte.
  • FIG. 1A shows a simplified schematic diagram of the front structure of a wound battery cell assembly
  • FIG. 1B shows a simplified schematic diagram of the top view structure of the battery cell assembly.
  • the cell assembly 100 includes a negative electrode piece 101 , a positive electrode piece 102 that are wound after being laminated, and a separator 103 located between the negative electrode piece 101 and the positive electrode piece 102 .
  • the negative electrode tab 101 includes a negative electrode tab body (not shown in the figure) provided with active material, and a plurality of negative electrode tabs 11 without an active material layer. The plurality of negative electrode tabs 11 are along the edges of the negative electrode tab body.
  • the edges on one side are arranged at intervals and stacked after being rolled to serve as the negative electrode of the battery core assembly 100 .
  • the positive electrode tab 102 includes a positive electrode tab body (not shown in the figure) provided with active material, and positive electrode tabs 12 without an active material layer.
  • the plurality of positive electrode tabs 12 are along one side of the positive electrode tab body. The edges are arranged at intervals and stacked after being rolled, thereby serving as the positive electrode of the cell assembly 100 and being spaced apart from the negative electrode.
  • detecting defects in the appearance of the tabs mainly refers to inspecting the appearance of multiple stacked negative electrode tabs and multiple stacked positive electrode tabs of the wound battery core assembly. Determine whether the overall dimensions of the stacked tabs meet the design requirements.
  • the inventor has conducted in-depth research and provided a method and device for detecting the appearance of tabs of battery core components, electronic equipment, computer-readable storage media and computer program products to improve the accuracy of appearance detection of tabs. properties, thereby improving the production yield rate of cell components and batteries.
  • the embodiment of the present disclosure uses computer vision technology to detect the appearance of the stacked tabs of the battery core assembly based on the captured image of the battery core assembly.
  • the root corner of the tab stacking area is determined in the detection image; then, the two sides of the main body area of the cell component image area are determined in the detection image; then, based on the two sides, during the detection
  • the reference edge line is determined in the image; then, the detection result information of the tab appearance can be obtained based on the relative positional relationship between the root corner point of the tab stacking area and the reference edge line.
  • the tab stacking area reflects the overall shape of multiple tabs after stacking. Each tab stacking area includes two root corner points. If the distance between the two root corner points is too large, it means that the tabs are stacked.
  • the battery disclosed in the embodiment of the present disclosure may be a power battery or an energy storage battery.
  • the application scenarios of power batteries include but are not limited to vehicles, ships, aircraft, spacecraft, electric tools, electric toys, various mobile terminals, etc.
  • the application scenarios of energy storage batteries include but are not limited to solar power generation systems, hydroelectric power generation systems, wind power generation systems, etc.
  • a method 200 for detecting the appearance of tabs of a battery core assembly includes the following steps S21 to S25.
  • a detection image 2000 including a background area 201 and a battery cell assembly image area 202 is obtained, wherein the battery cell assembly image area 202 includes a main body area 20 and a plurality of tab stacking areas (negative electrode tabs as shown in the figure). Stacking area 21 and positive electrode tab stacking area 22), each tab stacking area is adjacent to the top edge or bottom edge of the main body area 20.
  • each root corner point of the plurality of pole tab stacking areas in the detection image is determined (root corner points P1, P2, P3, P4 as shown in the figure).
  • step S23 two sides of the main body area 20 in the detection image (sides S1 and S2 as shown in the figure) are determined.
  • step S24 at least one reference edge line (reference edge lines L1, L2 as shown in the figure) is determined in the detection image based on two sides of the main body area 20 in the detection image.
  • step S25 based on each root corner point of the plurality of tab stacking areas in the detection image and at least one reference edge line, the detection result information of the tab appearance is determined (the detection result information includes, for example, the distance W1 shown in the figure, W2).
  • the cell assembly 100 is a wound cell assembly, including a main body 10 and a plurality of positive electrode tabs 12 located at the top of the main body 10 and arranged in a stack (as The positive electrode of the battery cell assembly), and a plurality of negative electrode tabs 11 (as the negative electrode of the battery cell assembly) located at the top of the main body 10 and arranged in a stack.
  • the detection image 2000 obtained by photographing the cell assembly 100 includes a background area 201 and a cell assembly image area 202.
  • the cell assembly image area 202 includes a positive electrode tab stacking area 22 and a negative electrode tab stack. Area 21, the positive electrode tab stacking area 22 and the negative electrode tab stacking area 21 are jointly adjacent to the top edge of the main body area 20 (according to the placement method of the cell assembly, it may also be adjacent to the bottom edge of the main body area) .
  • the battery core assembly is a wound battery core assembly, and its structure can also be designed to include a main body, a plurality of positive electrode tabs (as the positive electrodes of the battery core assembly) located at the top of the main body and arranged in a stack. , a plurality of negative electrode tabs (as the negative electrode of the battery cell assembly) located at the lower end of the main body and arranged in a stack.
  • the inspection image obtained by photographing the battery cell assembly includes a background area and a battery cell assembly image area.
  • the battery cell assembly image area includes a positive electrode tab stacking area and a negative electrode tab stacking area.
  • the positive electrode tab stacking area and the main body area The top edge is adjacent to (according to the placement of the battery cell components, it can also be adjacent to the bottom edge of the main body area), and the negative tab stacking area is adjacent to the bottom edge of the main body area (according to the placement of the battery core components) way, it can also be adjacent to the top edge of the main body area).
  • the method 200 of the embodiment of the present disclosure uses computer vision technology to detect the appearance of the stacked tabs of the battery core assembly based on the image of the battery core assembly. Compared with the manual measurement or detection methods with the help of special fixtures in the related art, Not only does it have high detection efficiency, it is applicable to a wide range of product specifications, but it can also significantly improve the accuracy of detection, thereby improving the production yield of cell components and batteries.
  • the detection result information of the tab appearance may include, for example: the distance between the two root corner points of each tab stacking area (ie, the distances W1, W2 shown in Figure 3), and/or the distance of the tab appearance. Pass or fail assessment result.
  • the distance between the two root corners of the tab stacking area exceeds the distance threshold, it can be determined that the appearance of the tab is unqualified, so that the defective battery cell assembly can be detected in time to prevent it from flowing into subsequent products. In the production process, this can improve the production yield of cell components and batteries.
  • the battery core assembly can also be evaluated for assembly based on the distance between two root corners of the tab stacking area. For example, when the distance between the two root corners of the pole tab stacking area exceeds the distance threshold, it means that the pole piece is wound loosely or too tightly, causing serious misalignment of the pole tabs after stacking. Product design and process can be carried out based on this. related review.
  • the detection image in step S21 can be obtained by photographing the battery cell assembly with an image acquisition device such as an industrial camera.
  • an image acquisition device such as an industrial camera.
  • the aperture of the image acquisition device can be adjusted to the maximum, and the shooting exposure time can be set to no less than 8000 microseconds and no more than 12000 microseconds, thereby obtaining
  • the high-exposure detection image makes the grayscale difference between the battery cell component image area and the background area obvious.
  • the image acquisition device and the battery core assembly can be accurately positioned before shooting, and then the image acquisition device can be started to shoot the battery core assembly, so that the top edge of the main body area in the image can be directly detected.
  • the bottom edge is approximately parallel to the row direction (i.e., the row direction of pixels in the image), and the two side edges are approximately orthogonal to the row direction (i.e., the column direction of pixels in the image).
  • the requirements for the shooting and positioning accuracy of the image acquisition device and the battery cell assembly are not high.
  • the detection image can be rotated and corrected.
  • the above-mentioned step S21 includes: performing rotation correction on the detected image so that the top and bottom edges of the main body area are substantially parallel to the row direction (ie, the row direction of pixels in the image), and the two side edges are parallel to the row direction. (i.e., the column directions of pixels in the image) are roughly orthogonal.
  • row direction and column direction are relative terms and do not represent absolute directions.
  • Approximately parallel and “approximately orthogonal” can be understood as being accepted within a certain error range. It is not required to be absolutely parallel or orthogonal.
  • step S25 the detection result information of the tab appearance is determined based on the relative positional relationship between the root corner point of the tab stacking area and the reference edge line in the detection image, and the detection result information is hardly affected by the detection of the main body area. Effect of relative position in the image.
  • the above-mentioned step S21 includes: performing binarization processing on the detection image 2000 to obtain a binary image 230, in which the battery cell component image area 202 serves as a binary image. 230 region of interest.
  • the binarization of the image is to set the gray value of the pixels on the image to 0 or 255 based on the set gray threshold, that is, the entire image presents an obvious visual effect of only black and white (in the figure 3, the grayscale value of the background area of the binary image 230 is 0, indicated by shading).
  • the battery cell component image area 202 serves as the area of interest.
  • the detection image is binarized to facilitate accurate extraction of the battery cell component image area 202 in subsequent steps.
  • the above step S22 includes: in the binary image 230, based on the template image 232 and the local search area 231 configured correspondingly for each root corner point, in each local search area 231 perform matching and positioning on the corresponding template image 232; and, based on the matching and positioning results of the template image 232 in each local search area 231, determine each root corner point of the multiple tab stacking areas in the binary image 230 (as shown in the figure) The root corner points P1, P2, P3, P4 shown in .
  • the template image 232 may be obtained based on the cell assembly image template.
  • the battery cell component image template refers to the standard image of the battery cell component used as a baseline reference.
  • the template image 232 is a part cut out from the cell assembly image template and includes the root corner point P', which can be used as a reference for image feature matching and comparison. To facilitate calculation, the root corner point P' in the template image 232 can be set at the center point of the template image.
  • the local search area 231 is a preset small-scale search area for the binary image 230. Matching and positioning the template image 232 in the local search area 231 can greatly reduce the amount of matching calculations and improve the speed and accuracy of the matching calculations.
  • the matching position of the template image 232 in the local search area 231 can be based on As a result (for example, the pixel coordinates of one corner point of the template image 232), through coordinate offset, the root corner point in the local search area 231 is determined, that is, the pixel coordinates of the root corner point in the local search area 231 are determined.
  • the above step S23 includes the following sub-steps one to four.
  • the battery cell component image area is extracted from the binary image.
  • the battery cell component image area can be extracted from the binary image based on at least one of a threshold segmentation algorithm, an image segmentation algorithm, and a maximum connected domain algorithm.
  • the battery component image area is rolled (i.e., translated) along the row direction in the binary image by one pixel column to obtain a contrast image.
  • sub-step three based on the grayscale difference between the binary image and the comparison image, determine multiple edge pixels in the battery cell component image area in the binary image (the grayscale of the edge pixels in the binary image and the comparison image The difference is 255 or -255).
  • edge pixels arranged in the first direction but with a distance T smaller than the distance threshold are filtered out from multiple edge pixels (pixels at points Q1 and Q2 in Figure 3, which may actually be arranged in the first direction). of multiple pairs of pixels) to obtain two sides of the main body area (which may be discontinuous), in which the first direction is approximately the same as the extension direction of the top and bottom edges of the main body area (accepted within a certain error range, that is, Can).
  • the distance threshold in the first direction can be determined based on the distance between two sides of the main body area in the aforementioned cell component image template. For example, it can be equal to or slightly smaller than the distance between the two sides.
  • it is possible to filter out edge pixels that are not related to the two sides of the main body area (such as edge pixels in the tab stack area), and on the other hand, it is also possible to try to avoid errors caused by poor hardware or shooting environment. The image features are missing or wrong, thereby further improving the accuracy of detection.
  • the above step S24 includes: performing straight line fitting (for example, through the Ransac algorithm) on the two sides S1 and S2 of the main body area 20 in the binary image 230 to obtain a straight line extension.
  • two fitting sides and, based on the two fitting sides, at least one reference edge line (reference edge lines L1, L2 as shown in the figure) is determined in the binary image.
  • the Ransac algorithm is a random parameter estimation iterative algorithm. Its basic principle is to first randomly select two points on one side, use this data set to calculate the data model, and then bring all the edge points on that side into the data model calculation. Find the number of all "inner points”, and through continuous iteration, compare the number of "inner points” between the current model and the best model previously introduced, record the model parameters and the number of "inner points” with the largest number of "inner points", When "the number of interior points is greater than a certain number", the final fitted straight line parameters are obtained.
  • the distance between the two root corner points of each tab stacking area is calculated distance W1, W2, and then determine whether the appearance of the tabs of the battery cell assembly is qualified, and the detection accuracy is high.
  • the two fitting sides may be used as the first reference edge line and the second reference edge line respectively, based on the determined two root corner points of the positive electrode tab stacking area and the adjacent first
  • the distance between the two root corner points of the positive electrode tab stacking area is calculated, based on the determined relative distance between the two root corner points of the negative electrode tab stacking area and the adjacent second reference edge line.
  • distance calculate the distance between the two root corners of the negative electrode tab stacking area, and then determine whether the appearance of the tabs of the battery cell assembly is qualified.
  • some embodiments of the present disclosure provide a method for detecting the appearance of tabs of a battery core assembly, including the following steps S401 to S410.
  • step S401 a detection image collected by the image collection device is obtained.
  • step S402 the detection image is binarized to obtain a binarized image with the battery cell component image area as the area of interest.
  • step S403 in the binary image, based on the template image and local search area respectively configured for each root corner point of the positive electrode tab stacking area and the negative electrode tab stacking area, the corresponding template image is performed in each local search area. Match positioning search.
  • each root corner point in the binary image is determined based on the matching positioning result of the template image in each local search area.
  • step S405 the battery cell component image area is extracted from the binary image based on at least one of a threshold segmentation algorithm, an image segmentation algorithm, and a maximum connected domain algorithm.
  • step S406 the battery component image area is scrolled by one pixel column along the row direction in the binary image to obtain a comparison image.
  • step S407 multiple edge pixels of the battery cell component image area in the binary image are determined based on the grayscale difference between the binary image and the comparison image.
  • step S408 edge pixels arranged in the first direction but with a distance smaller than the distance threshold are filtered out from the plurality of edge pixels to obtain two sides of the main body area, where the first direction is consistent with the top or bottom edge of the main body area.
  • the extension directions are roughly the same.
  • step S409 perform straight line fitting on the two sides of the main body area in the binary image, respectively, to obtain two straight-line fitted sides, which are used as the first reference edge line and the second reference edge line respectively.
  • step S410 based on the distances from the two root corners of the positive electrode tab stacking area to the first reference edge line and the distances from the two root corners of the negative electrode tab stacking area to the second reference edge line, the battery core is determined. Whether the appearance of the module’s tabs is acceptable.
  • an embodiment of the present disclosure also provides a device 500 for detecting the appearance of tabs of a battery core assembly, including: an acquisition unit 501 configured to acquire a detection image including a background area and an image area of the battery core assembly, wherein, The battery core component image area includes a main body area and a plurality of tab stacking areas, each tab stacking area being adjacent to the top or bottom edge of the main body area; the first determination unit 502 is configured to determine multiple poles in the detection image.
  • each root corner point of the ear stacking area is configured to determine two sides of the main body area in the detection image; the third determination unit 504 is configured to determine based on the two sides of the main body area in the detection image, At least one reference edge line is determined in the detection image; and, the fourth determination unit 505 is configured to determine the tab appearance based on each root corner point of the plurality of tab stacking areas in the detection image and the at least one reference edge line. test result information.
  • An embodiment of the present disclosure also provides an electronic device, including 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 at least one processor, and the instructions are executed by at least one processor, To enable at least one processor to perform the steps of the aforementioned method.
  • the appearance of stacked tabs of the battery cell assembly can be inspected with high detection accuracy, thereby improving the production yield of the cell assembly and the battery.
  • Embodiments of the present disclosure also provide a computer-readable storage medium storing computer instructions, wherein the computer instructions are configured to cause the computer to execute the method of any of the foregoing embodiments.
  • An embodiment of the present disclosure also provides a computer program product, including a computer program, wherein the computer program implements the method of any of the foregoing embodiments when executed by a processor.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Chemical & Material Sciences (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Electrochemistry (AREA)
  • General Chemical & Material Sciences (AREA)
  • Quality & Reliability (AREA)
  • Manufacturing & Machinery (AREA)
  • Image Analysis (AREA)
  • Secondary Cells (AREA)

Abstract

提供一种检测电芯组件的极耳外观的方法与装置、电子设备、计算机可读存储介质及计算机程序产品。方法包括：获取包括背景区和电芯组件图像区的检测图像，其中，电芯组件图像区包括主体区和多个极耳堆叠区，每个极耳堆叠区与主体区的顶边或底边相邻接；确定检测图像中多个极耳堆叠区的每个根部角点；确定检测图像中主体区的两个侧边；基于检测图像中主体区的两个侧边，在检测图像中确定出至少一个参照边缘线；以及，基于检测图像中多个极耳堆叠区的每个根部角点、以及至少一个参照边缘线，确定极耳外观的检测结果信息。

Description

检测电芯组件的极耳外观的方法与装置、电子设备 技术领域
本公开涉及电池生产技术领域，尤其涉及一种检测电芯组件的极耳外观的方法与装置、电子设备、计算机可读存储介质及计算机程序产品。
背景技术
在相关技术中，充电电池(指在电池放电后可通过充电的方式使活性物质激活而继续使用的电池，又称二次电池，以下简称为电池)包括电池箱以及位于电池箱内的通过串联和/或并联方式组合的多个电池单体。电池单体是电池中提供能量来源的最小单元，而电芯组件是电池单体中发生电化学反应的关键部件，对电芯组件的极耳外观进行缺陷检测，是电池生产中的重要一环。
发明内容
本公开旨在至少解决相关技术中存在的技术问题之一。为此，本公开的一个目的在于提出一种检测电芯组件的极耳外观的方法与装置、电子设备、计算机可读存储介质及计算机程序产品。
本公开第一方面的实施例提供了一种检测电芯组件的极耳外观的方法，包括：获取包括背景区和电芯组件图像区的检测图像，其中，电芯组件图像区包括主体区和多个极耳堆叠区，每个极耳堆叠区与主体区的顶边或底边相邻接；确定检测图像中多个极耳堆叠区的每个根部角点；确定检测图像中主体区的两个侧边；基于检测图像中主体区的两个侧边，在检测图像中确定出至少一个参照边缘线；以及，基于检测图像中多个极耳堆叠区的每个根部角点、以及至少一个参照边缘线，确定极耳外观的检测结果信息。
本公开实施例的方法，基于电芯组件的图像，利用计算机视觉技术，对电芯组件的堆叠极耳的外观进行检测，相比相关技术中人工测量或者借助于专用治具的检测方式，不但检测效率较高，适用产品规格广泛，而且可以显著提高检测的准确性，进而提高电芯组件以及电池的生产良品率。
在一些实施例中，获取检测图像，包括：对检测图像进行二值化处理，得到二值化图像，其中，电芯组件图像区作为二值化图像的感兴趣区域。对检测图像进行二值化处理，便于在后续步骤中准确提取电芯组件图像区。
在一些实施例中，确定检测图像中多个极耳堆叠区的每个根部角点，包括：在二值化图像中，基于为每个根部角点对应配置的模板图像和局部搜索区域，在每个局部搜索区域对相应模板图像进行匹配定位；以及，基于每个局部搜索区域内模板图像的匹配定位结果，确定二值化图像中多个极耳堆叠区的每个根部角点。
在一些实施例中，模板图像基于电芯组件图像模板获得。
局部搜索区域是针对二值化图像预先设定的小范围搜索区域，在局部搜索区域对模板图像进行匹配定位，可以大大减少匹配计算量，提高匹配计算的速度和准确性。
在一些实施例中，确定检测图像中主体区的两个侧边，包括：从二值化图像中提取电芯组件图像区；将电芯组件图像区在二值化图像中沿行向滚动一个像素列，得到对比图像；基于二值化图像和对比图像的灰度差值，确定二值化图像中电芯组件图像区的多个边缘像素；以及，从多个边缘像素中滤除在第一方向上排列但间距小于距离阈值的边缘像素，得到二值化图像中主体区的两个侧边，其中，第一方向与主体区的顶边和底边的延伸方向大致相同。
采用该实施例方案，一方面可以过滤掉与主体区的两个侧边不相关的边缘像素，另一方面还可以尽量避免因硬件或者拍摄环境不佳导致的图像特征缺失或错误，从而进一步提高检测的准确性。
在一些实施例中，基于检测图像中主体区的两个侧边，在检测图像中确定出至少一个参照边缘线，包括：对二值化图像中主体区的两个侧边分别进行直线拟合，得到直线延伸的两个拟合侧边；以及，基于两个拟合侧边，在二值化图像中确定出至少一个参照边缘线。基于直线拟合得到的拟合侧边确定参照边缘线，可以使得检测的准确性进一步提高。
在一些实施例中，从二值化图像中提取电芯组件图像区，包括：基于阈值分割算法、图像分割算法和最大连通域算法中的至少一种，从二值化图像中提取电芯组件图像区。
在一些实施例中，获取检测图像，包括：对检测图像进行旋转校正，以使主体区的顶边和底边大致与行向平行，以及使主体区的两个侧边大致与行向正交。
在一些实施例中，检测图像的拍摄曝光时间不小于8000微秒，且不大于12000微秒。获得高曝光度的检测图像，使电芯组件图像区与背景区的灰度明显区别，有利于进一步提高检测的准确性。
在一些实施例中，极耳外观的检测结果信息包括：每个极耳堆叠区的两个根部角点之间的距离；和/或，极耳外观的合格或不合格评估结果。
在一些实施例中，电芯组件图像区包括两个极耳堆叠区，分别为正极极耳堆叠区和负极极耳堆叠区，其中，正极极耳堆叠区和负极极耳堆叠区共同与主体区的顶边或底边相邻接，或者，正极极耳堆叠区与主体区的顶边相邻接，负极极耳堆叠区与主体区的底边相邻接。
本公开第二方面的实施例提供了一种检测电芯组件的极耳外观的装置，包括：获取单元，配置为获取包括背景区和电芯组件图像区的检测图像，其中，电芯组件图像区包括主体区和多个极耳堆叠区，每个极耳堆叠区与主体区的顶边或底边相邻接；第一确定单元，配置为确定检测图像中多个极耳堆叠区的每个根部角点；第二确定单元，配置为确定检测图像中主体区的两个侧边；第三确定单元，配置为基于检测图像中主体区的两个侧边，在检测图像中确定出至少一个参照边缘线；以及，第四确定单元，配置为基于检测图像中多个极耳堆叠区的每个根部角点、以及至少一个参照边缘线，确定极耳外观的检测结果信息。
本公开第三方面的实施例提供了一种存储有计算机指令的计算机可读存储介质，计算机指令配置为使计算机执行前述方面的方法。
本公开第三方面的实施例提供了一种计算机程序产品，包括计算机程序，计算机程序在被处理器执行时实现前述方面的方法。
采用本公开以上实施例对电芯组件的堆叠极耳的外观进行检测，不但检测效率较高，适用产品规格广泛，而且可以显著提高检测的准确性，进而提高电芯组件以及电池的生产良品率。
上述说明仅是本公开技术方案的概述，为了能够更清楚了解本公开的技术手段，而可依照说明书的内容予以实施，并且为了让本公开的上述和其它目的、特征和优点能够更明显易懂，以下特举本公开的具体实施方式。
附图说明
在附图中，除非另外规定，否则贯穿多个附图相同的附图标记表示相同或相似的部件或元素。这些附图不一定是按照比例绘制的。应该理解，这些附图仅描绘了根据本公开公开的一些实施方式，而不应将其视为是对本公开范围的限制。
图1A为一种卷绕式电芯组件的主视结构简化示意图；
图1B为一种卷绕式电芯组件的俯视结构简化示意图；
图2为本公开一些实施例的检测电芯组件的极耳外观的方法的流程示意图；
图3为本公开一些实施例的检测电芯组件的极耳外观的方法的流程示意图；
图4为本公开一些实施例的检测电芯组件的极耳外观的方法的流程示意图；
以及，图5为本公开一些实施例的检测电芯组件的极耳外观的装置的结构框图。
附图标记说明：
100-电芯组件
10-主体部
101-负极极片
11-负极极耳
102-正极极片
12-正极极耳
103-隔膜
2000-检测图像
201-背景区
202-电芯组件图像区
20-主体区
21-负极堆叠极耳区
22-正极堆叠极耳区
230-二值化图像
231-局部搜索区域
232-模板图像
具体实施方式
下面将结合附图对本公开技术方案的实施例进行详细的描述。以下实施例仅用于更加清楚地说明本公开的技术方案，因此只作为示例，而不能以此来限制本公开的保护范围。
除非另有定义，本文所使用的所有的技术和科学术语与属于本公开的技术领域的技术人员通常理解的含义相同；本文中所使用的术语只是为了描述具体的实施例的目的， 不是旨在于限制本公开；本公开的说明书和权利要求书及上述附图说明中的术语“包括”和“具有”以及它们的任何变形，意图在于覆盖不排他的包含。
在本公开实施例的描述中，技术术语“第一”“第二”等仅用于区别不同对象，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量、特定顺序或主次关系。在本公开实施例的描述中，“多个”的含义是两个以上，除非另有明确具体的限定。
在本文中提及“实施例”意味着，结合实施例描述的特定特征、结构或特性可以包含在本公开的至少一个实施例中。在说明书中的各个位置出现该短语并不一定均是指相同的实施例，也不是与其它实施例互斥的独立的或备选的实施例。本领域技术人员显式地和隐式地理解的是，本文所描述的实施例可以与其它实施例相结合。
在本公开实施例的描述中，术语“和/或”仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中字符“/”，一般表示前后关联对象是一种“或”的关系。
在本公开实施例的描述中，术语“多个”指的是两个以上(包括两个)，同理，“多组”指的是两组以上(包括两组)，“多片”指的是两片以上(包括两片)。
在本公开实施例的描述中，技术术语“中心”、“列向”、“行向”、“长度”、“宽度”、“厚度”、“上”、“下”、“前”、“后”、“左”、“右”、“竖直”、“水平”、“顶”、“底”、“内”、“外”、“顺时针”、“逆时针”、“轴向”、“径向”、“周向”等指示的方位或位置关系为基于附图所示的方位或位置关系，仅是为了便于描述本公开实施例和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本公开实施例的限制。
在本公开实施例的描述中，除非另有明确的规定和限定，技术术语“安装”、“相连”“连接”、“固定”等术语应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或成一体；也可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连，可以是两个元件内部的连通或两个元件的相互作用关系。对于本领域的普通技术人员而言，可以根据具体情况理解上述术语在本公开实施例中的具体含义。
常见的电芯组件主要包括叠片式和卷绕式两种。叠片式电芯组件包括交替层叠的多个正极极片和多个负极极片，以及在任意相邻的正极极片和负极极片之间设置的隔膜。卷绕式电芯组件包括在层叠后卷绕的正极极片、负极极片以及位于正极极片和负极极片之间的隔膜。以锂离子电池单体为例，在电池单体的壳体内部，电芯组件被电解液浸润，锂离子以电解液为介质在负极和正极之间运动，从而可以使电池单体实现充电与放电， 隔膜的作用是允许锂离子自由通过，而不允许电子通过，从而防止电池单体的负极和正极之间通过电解液发生短路。
图1A所示为一种卷绕式电芯组件的主视结构简化示意图，图1B所示为该电芯组件的俯视结构简化示意图。如图1A和图1B所示，该电芯组件100包括在层叠后卷绕的负极极片101、正极极片102以及位于负极极片101和正极极片102之间的隔膜103。负极极片101包括设有活性物质的负极极片本体(图中未示出)，以及未设有活性物质层的多个负极极耳11，该多个负极极耳11沿负极极片本体的一侧边缘间隔排列，在卷绕后堆叠，从而作为电芯组件100的负极。正极极片102包括设有活性物质的正极极片本体(图中未示出)，以及未设有活性物质层的正极极耳12，该多个正极极耳12沿正极极片本体的一侧边缘间隔排列，在卷绕后堆叠，从而作为电芯组件100的正极并且与负极相间隔。
电芯组件在生产过程中，由于环境、工艺及设备原因，可能造成各种缺陷，因此，需要对电芯组件进行缺陷检测，以保证一定的出厂良品率，其中，对极耳外观进行缺陷检测是极其重要的一环，其检测结果的有效性直接影响到电池出厂的安全性。在本公开实施例中，对极耳外观进行缺陷检测主要是指：对卷绕式电芯组件的相堆叠的多个负极极耳和相堆叠的多个正极极耳进行外观检测，以此来判断堆叠后的极耳整体尺寸是否符合设计要求。
基于以上考虑，发明人经过深入研究，提供了一种检测电芯组件的极耳外观的方法与装置、电子设备、计算机可读存储介质及计算机程序产品，以提高对极耳进行外观检测的准确性，进而提高电芯组件以及电池的生产良品率。
本公开实施例方案基于拍摄的电芯组件的图像，利用计算机视觉技术，对电芯组件的堆叠极耳的外观进行检测。首先，在检测图像中确定出极耳堆叠区的根部角点；然后，在检测图像中确定出电芯组件图像区的主体区的两个侧边；然后，基于该两个侧边，在检测图像中确定出参照边缘线；之后，可以基于极耳堆叠区的根部角点和参照边缘线之间的相对位置关系，得到极耳外观的检测结果信息。极耳堆叠区反映了多个极耳在堆叠后的整体外形，每个极耳堆叠区包括两个根部角点，如果该两个根部角点之间的距离过大则说明极耳在堆叠后错位严重，需要进行产品设计和工艺上的相关检讨。采用本公开实施例技术方案，可以在电芯组件制作完成后，对其堆叠极耳进行外观检测，可以及时发现相关缺陷，而且检测的准确性较高，从而可以提高电芯组件以及电池的生产良品率。
本公开实施例中公开的电池可以是动力电池或储能电池。其中，动力电池的应用场景包括但不限于车辆、船舶、飞行器、航天器、电动工具、电动玩具，各类移动终端等等。储能电池的应用场景包括但不限于太阳能发电系统、水力发电系统、风力发电系统，等等。
如图2和图3所示，本公开实施例提供的检测电芯组件的极耳外观的方法200，包括以下步骤S21至步骤S25。
在步骤S21，获取包括背景区201和电芯组件图像区202的检测图像2000，其中，电芯组件图像区202包括主体区20和多个极耳堆叠区(如图中所示的负极极耳堆叠区21和正极极耳堆叠区22)，每个极耳堆叠区与主体区20的顶边或底边相邻接。
在步骤S22，确定检测图像中多个极耳堆叠区的每个根部角点(如图中所示的根部角点P1,P2,P3,P4)。
在步骤S23，确定检测图像中主体区20的两个侧边(如图中所示的侧边S1,S2)。
在步骤S24，基于检测图像中主体区20的两个侧边，在检测图像中确定出至少一个参照边缘线(如图中所示的参照边缘线L1,L2)。
在步骤S25，基于检测图像中多个极耳堆叠区的每个根部角点、以及至少一个参照边缘线，确定极耳外观的检测结果信息(检测结果信息例如包括图中所示的距离W1,W2)。
本公开实施例可以用于多种电芯组件产品的堆叠极耳的外观检测。如图1A和图1B所示，在一些实施例中，电芯组件100为卷绕式电芯组件，包括主体部10、位于主体部10的顶端且堆叠设置的多个正极极耳12(作为电芯组件的正极)，位于主体部10的顶端且堆叠设置的多个负极极耳11(作为电芯组件的负极)。如图3所示，对该电芯组件100进行拍摄获得的检测图像2000，包括背景区201和电芯组件图像区202，电芯组件图像区202包括正极极耳堆叠区22和负极极耳堆叠区21，该正极极耳堆叠区22和负极极耳堆叠区21共同与主体区20的顶边相邻接(根据电芯组件的放置方式，也可以是与主体区的底边相邻接)。
在另一些实施例中，电芯组件为卷绕式电芯组件，其结构也可以设计为包括主体部、位于主体部的顶端且堆叠设置的多个正极极耳(作为电芯组件的正极)，位于主体部的低端且堆叠设置的多个负极极耳(作为电芯组件的负极)。对该电芯组件进行拍摄获得的检测图像，包括背景区和电芯组件图像区，电芯组件图像区包括正极极耳堆叠区和负极极耳堆叠区，其中，正极极耳堆叠区与主体区的顶边相邻接(根据电芯组件的放置 方式，也可以是与主体区的底边相邻接)，负极极耳堆叠区与主体区的底边相邻接(根据电芯组件的放置方式，也可以是与主体区的顶边相邻接)。
本公开实施例的方法200，基于电芯组件的图像，利用计算机视觉技术，对电芯组件的堆叠极耳的外观进行检测，相比相关技术中人工测量或者借助于专用治具的检测方式，不但检测效率较高，适用产品规格广泛，而且可以显著提高检测的准确性，进而提高电芯组件以及电池的生产良品率。
其中，极耳外观的检测结果信息例如可以包括：每个极耳堆叠区的两个根部角点之间的距离(即图3中所示的距离W1,W2)，和/或极耳外观的合格或不合格评估结果。
在一些实施例中，当极耳堆叠区的两个根部角点之间的距离超过距离阈值时，可以判定极耳外观不合格，从而及时将该电芯组件不良品检出，避免其流入后续生产工序中，以此提高电芯组件以及电池的生产良品率。
在一些实施例中，还可以基于极耳堆叠区的两个根部角点之间的距离，对电芯组件进行组装性评价。例如，当极耳堆叠区的两个根部角点之间的距离超过距离阈值时，说明极片卷绕过松或过紧导致极耳在堆叠后错位严重，可以基于此进行产品设计和工艺上的相关检讨。
步骤S21中的检测图像可以由工业相机等图像采集设备对电芯组件进行拍摄获得。在一些实施例中，当拍摄环境明亮度一般或者较暗时，可以将图像采集设备的光圈尽量调至最大，将拍摄曝光时间设置为不小于8000微秒，且不大于12000微秒，从而获得高曝光度的检测图像，使电芯组件图像区与背景区的灰度明显区别。
在本公开的一些实施例中，可以在拍摄前先将图像采集设备与电芯组件精确定位，然后再启动图像采集设备对电芯组件进行拍摄，从而可以直接使检测图像中主体区的顶边和底边与行向(即图像中的像素行向)大致平行，两个侧边与行向(即图像中的像素列向)大致正交。
在本公开的另一些实施例中，对图像采集设备与电芯组件的拍摄定位精度要求不高，可以在拍摄到检测图像后，对检测图像进行旋转校正。在该实施例中，上述步骤S21包括：对检测图像进行旋转校正，以使主体区的顶边和底边与行向(即图像中的像素行向)大致平行，两个侧边与行向(即图像中的像素列向)大致正交。
在本公开实施例中，“行向”和“列向”是相对而言的，不表示绝对方向，“大致平行”、“大致正交”可以理解为在一定误差范围内被接受即可，并不要求其绝对平行或者正交。
需要说明的是，在本公开实施例中，即使主体区的顶边和底边相对行向不平行，上述方法200仍然适用，而且仍能获得较为准确的检测结果信息。这是因为，在步骤S25，是基于检测图像中极耳堆叠区的根部角点和参照边缘线的相对位置关系来确定极耳外观的检测结果信息的，检测结果信息几乎不受主体区在检测图像中相对位置的影响。
在本公开的一些实施例中，如图3所示，上述步骤S21包括：对检测图像2000进行二值化处理，得到二值化图像230，其中，电芯组件图像区202作为二值化图像230的感兴趣区域。
图像的二值化，就是基于设定的灰度阈值，将图像上的像素点的灰度值设置为0或255，也就是将整个图像呈现出明显的只有黑和白的视觉效果(在图3中，二值化图像230的背景区的灰度值为0，以阴影示意)。经过二值化处理得到二值化图像230中，电芯组件图像区202作为感兴趣区域。对检测图像进行二值化处理，便于在后续步骤中准确提取电芯组件图像区202。
在一些实施例中，如图3所示，上述步骤S22包括：在二值化图像230中，基于为每个根部角点对应配置的模板图像232和局部搜索区域231，在每个局部搜索区域231对相应模板图像232进行匹配定位；以及，基于每个局部搜索区域231内模板图像232的匹配定位结果，确定二值化图像230中多个极耳堆叠区的每个根部角点(如图中所示的根部角点P1,P2,P3,P4)。
模板图像232可以基于电芯组件图像模板获得。电芯组件图像模板是指用作基准参照的电芯组件的标准图像。模板图像232是从电芯组件图像模板中截取的一部分并且包含了根部角点P’，可以用作图像特征匹配比对的参照。为了方便计算，可以将模板图像232中根部角点P’设在该模板图像的中心点处。
局部搜索区域231是针对二值化图像230预先设定的小范围搜索区域，在局部搜索区域231对模板图像232进行匹配定位，可以大大减少匹配计算量，提高匹配计算的速度和准确性。
在局部搜索区域231内搜索到模板图像232对应的匹配图像(例如模板图像和匹配图像的特征相似度大于设定的相似度阈值)之后，可以基于模板图像232在局部搜索区域231内的匹配定位结果(例如模板图像232的其中一个角点的像素坐标)，通过坐标坐标偏移，确定出局部搜索区域231内的根部角点，也即确定出局部搜索区域231内根部角点的像素坐标。
在一些实施例中，上述步骤S23包括以下子步骤一至子步骤四。
在子步骤一，从二值化图像中提取电芯组件图像区。例如，可以基于阈值分割算法、图像分割算法和最大连通域算法中的至少一种，从二值化图像中提取电芯组件图像区。
在子步骤二，将电芯组件图像区在二值化图像中沿行向滚动(即平移)一个像素列，得到对比图像。
在子步骤三，基于二值化图像和对比图像的灰度差值，确定二值化图像中电芯组件图像区的多个边缘像素(边缘像素在二值化图像和对比图像中的灰度相差255或-255)。
在子步骤四，从多个边缘像素中滤除在第一方向上排列但间距T小于距离阈值的边缘像素(如图3中Q1,Q2点处像素，实际上可以是在第一方向上排列的多对像素)，得到主体区的两个侧边(可以是不连续的)，其中，第一方向与主体区的顶边和底边的延伸方向大致相同(在一定误差范围内被接受即可)。
第一方向上的距离阈值可以基于前述电芯组件图像模板中主体区的两个侧边的间距来确定，例如可以等于或者略小于两个侧边的间距。采用该实施例方案，一方面可以过滤掉与主体区的两个侧边不相关的边缘像素(例如极耳堆叠区的边缘像素)，另一方面还可以尽量避免因硬件或者拍摄环境不佳导致的图像特征缺失或错误，从而进一步提高检测的准确性。
如图3所示，在一些实施例中，上述步骤S24包括：对二值化图像230中主体区20的两个侧边S1,S2分别进行直线拟合(例如通过Ransac算法)，得到直线延伸的两个拟合侧边；以及，基于两个拟合侧边，在二值化图像中确定出至少一个参照边缘线(如图中所示的参照边缘线L1,L2)。
Ransac算法是一种随机参数估计迭代算法，其基本原理为，首先随机在一个侧边选择两个点，使用这个数据集来计算出数据模型，再将所有的该侧边缘点带入数据模型计算出所有“内点”的数目，通过不断地迭代，比较当前模型和之前推出的最好的模型的“内点“的数量，记录最大“内点”数的模型参数和“内点”数，当“内点数目大于一定数量”时就得到了最终拟合的直线参数。
可以选用任意一个拟合侧边作为参照边缘线，也可以基于任意一个拟合侧边进行平行偏移，从而得到参照边缘线。基于在步骤S22确定的各个根部角点A1,,A2,A3,A4与该参照边缘线(如参照边缘线L1)的相对距离，计算出每个极耳堆叠区的两个根部角点之间的距离W1,W2，并继而判断电芯组件的极耳外观是否合格，检测的准确性较高。
在一些实施例中，也可以是，两个拟合侧边分别作为第一参照边缘线和第二参照边缘线，基于确定的正极极耳堆叠区的两个根部角点与相邻的第一参照边缘线的相对距离，计算出正极极耳堆叠区的两个根部角点之间的距离，基于确定的负极极耳堆叠区的两个根部角点与相邻的第二参照边缘线的相对距离，计算出负极极耳堆叠区的两个根部角点之间的距离，并继而判断电芯组件的极耳外观是否合格。
如图4所示，本公开一些实施例提供的检测电芯组件的极耳外观的方法，包括以下步骤S401至步骤S410。
在步骤S401，获取由图像采集设备采集的检测图像。
在步骤S402，对检测图像进行二值化处理，得到以电芯组件图像区作为感兴趣区域的二值化图像。
在步骤S403，在二值化图像中，基于为正极极耳堆叠区和负极极耳堆叠区的各个根部角点分别配置的模板图像和局部搜索区域，在每个局部搜索区域对相应模板图像进行匹配定位搜索。
在步骤S404，基于每个局部搜索区域内模板图像的匹配定位结果，确定二值化图像中的各个根部角点。
在步骤S405，基于阈值分割算法、图像分割算法和最大连通域算法中的至少一种，从二值化图像中提取电芯组件图像区。
在步骤S406，将电芯组件图像区在二值化图像中沿行向滚动一个像素列，得到对比图像。
在步骤S407，基于二值化图像和对比图像的灰度差值，确定二值化图像中电芯组件图像区的多个边缘像素。
在步骤S408，从多个边缘像素中滤除在第一方向上排列但间距小于距离阈值的边缘像素，得到主体区的两个侧边，其中，第一方向与主体区的顶边或底边的延伸方向大致相同。
在步骤S409，对二值化图像中主体区的两个侧边分别进行直线拟合，得到直线延伸的两个拟合侧边，分别作为第一参照边缘线和第二参照边缘线。
在步骤S410，基于正极极耳堆叠区的两个根部角点分别到第一参照边缘线的距离、负极极耳堆叠区的两个根部角点分别到第二参照边缘线的距离，判断电芯组件的极耳的外观是否合格。
如图5所示，本公开实施例还提供一种检测电芯组件的极耳外观的装置500，包括：获取单元501，配置为获取包括背景区和电芯组件图像区的检测图像，其中，电芯组件图像区包括主体区和多个极耳堆叠区，每个极耳堆叠区与主体区的顶边或底边相邻接；第一确定单元502，配置为确定检测图像中多个极耳堆叠区的每个根部角点；第二确定单元503，配置为确定检测图像中主体区的两个侧边；第三确定单元504，配置为基于检测图像中主体区的两个侧边，在检测图像中确定出至少一个参照边缘线；以及，第四确定单元505，配置为基于检测图像中多个极耳堆叠区的每个根部角点、以及至少一个参照边缘线，确定极耳外观的检测结果信息。
本公开实施例还提供一种电子设备，包括至少一个处理器以及与至少一个处理器通信连接的存储器，其中，存储器存储有可被至少一个处理器执行的指令，指令被至少一个处理器执行，以使至少一个处理器能够执行前述方法的步骤。
采用本公开上述实施例的装置和电子设备，可以对电芯组件的堆叠极耳进行外观检测，检测的准确性较高，从而可以提高电芯组件以及电池的生产良品率。
本公开实施例还提供一种存储有计算机指令的计算机可读存储介质，其中，计算机指令配置为使计算机执行前述任一实施例的方法。
本公开实施例还提供一种计算机程序产品，包括计算机程序，其中，计算机程序在被处理器执行时实现前述任一实施例的方法。
最后应说明的是：以上各实施例仅用以说明本公开的技术方案，而非对其限制；尽管参照前述各实施例对本公开进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本公开各实施例技术方案的范围，其均应涵盖在本公开的权利要求和说明书的范围当中。尤其是，只要不存在结构冲突，各个实施例中所提到的各项技术特征均可以任意方式组合起来。本公开并不局限于文中公开的特定实施例，而是包括落入权利要求的范围内的所有技术方案。

Claims (14)

  1. 一种检测电芯组件的极耳外观的方法，包括：
    获取包括背景区和电芯组件图像区的检测图像，其中，电芯组件图像区包括主体区和多个极耳堆叠区，每个极耳堆叠区与主体区的顶边或底边相邻接；
    确定检测图像中多个极耳堆叠区的每个根部角点；
    确定检测图像中主体区的两个侧边；
    基于检测图像中主体区的两个侧边，在检测图像中确定出至少一个参照边缘线；以及
    基于检测图像中多个极耳堆叠区的每个根部角点、以及至少一个参照边缘线，确定极耳外观的检测结果信息。
  2. 根据权利要求1所述的方法，其中，获取检测图像，包括：
    对检测图像进行二值化处理，得到二值化图像，其中，电芯组件图像区作为二值化图像的感兴趣区域。
  3. 根据权利要求2所述的方法，其中，确定检测图像中多个极耳堆叠区的每个根部角点，包括：
    在二值化图像中，基于为每个根部角点对应配置的模板图像和局部搜索区域，在每个局部搜索区域对相应模板图像进行匹配定位；以及
    基于每个局部搜索区域内模板图像的匹配定位结果，确定二值化图像中多个极耳堆叠区的每个根部角点。
  4. 根据权利要求3所述的方法，其中，模板图像基于电芯组件图像模板获得。
  5. 根据权利要求2所述的方法，其中，确定检测图像中主体区的两个侧边，包括：
    从二值化图像中提取电芯组件图像区；
    将电芯组件图像区在二值化图像中沿行向滚动一个像素列，得到对比图像；
    基于二值化图像和对比图像的灰度差值，确定二值化图像中电芯组件图像区的多个边缘像素；以及
    从多个边缘像素中滤除在第一方向上排列但间距小于距离阈值的边缘像素，得到二值化图像中主体区的两个侧边，其中，第一方向与主体区的顶边和底边的延伸方向大致相同。
  6. 根据权利要求5所述的方法，其中，基于检测图像中主体区的两个侧边，在检测图像中确定出至少一个参照边缘线，包括：
    对二值化图像中主体区的两个侧边分别进行直线拟合，得到直线延伸的两个拟合侧边；以及
    基于两个拟合侧边，在二值化图像中确定出至少一个参照边缘线。
  7. 根据权利要求5所述的方法，其中，从二值化图像中提取电芯组件图像区，包括：
    基于阈值分割算法、图像分割算法和最大连通域算法中的至少一种，从二值化图像中提取电芯组件图像区。
  8. 根据权利要求1所述的方法，其中，获取检测图像，包括：
    对检测图像进行旋转校正，以使主体区的顶边和底边大致与行向平行，以及使主体区的两个侧边大致与行向正交。
  9. 根据权利要求1所述的方法，其中，检测图像的拍摄曝光时间不小于8000微秒，且不大于12000微秒。
  10. 根据权利要求1所述的方法，其中，极耳外观的检测结果信息包括：
    每个极耳堆叠区的两个根部角点之间的距离；和/或
    极耳外观的合格或不合格评估结果。
  11. 根据权利要求1至10中任一项所述的方法，其中，
    电芯组件图像区包括两个极耳堆叠区，分别为正极极耳堆叠区和负极极耳堆叠区，其中，正极极耳堆叠区和负极极耳堆叠区共同与主体区的顶边或底边相邻接，或者，正极极耳堆叠区与主体区的顶边相邻接，负极极耳堆叠区与主体区的底边相邻接。
  12. 一种检测电芯组件的极耳外观的装置，包括：
    获取单元，配置为获取包括背景区和电芯组件图像区的检测图像，其中，电芯组件图像区包括主体区和多个极耳堆叠区，每个极耳堆叠区与主体区的顶边或底边相邻接；
    第一确定单元，配置为确定检测图像中多个极耳堆叠区的每个根部角点；
    第二确定单元，配置为确定检测图像中主体区的两个侧边；
    第三确定单元，配置为基于检测图像中主体区的两个侧边，在检测图像中确定出至少一个参照边缘线；以及
    第四确定单元，配置为基于检测图像中多个极耳堆叠区的每个根部角点、以及至少一个参照边缘线，确定极耳外观的检测结果信息。
  13. 一种存储有计算机指令的计算机可读存储介质，所述计算机指令配置为使计算机执行如权利要求1至11中任一项所述的方法。
  14. 一种计算机程序产品，包括计算机程序，所述计算机程序在被处理器执行时实现如权利要求1至11中任一项所述的方法。
PCT/CN2022/107073 2022-07-21 2022-07-21 检测电芯组件的极耳外观的方法与装置、电子设备 WO2024016266A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
PCT/CN2022/107073 WO2024016266A1 (zh) 2022-07-21 2022-07-21 检测电芯组件的极耳外观的方法与装置、电子设备
EP22888600.8A EP4339887A1 (en) 2022-07-21 2022-07-21 Method and apparatus for inspecting tab appearance of battery cell assembly, and electronic device
CN202280006741.3A CN117751381A (zh) 2022-07-21 2022-07-21 检测电芯组件的极耳外观的方法与装置、电子设备
US18/138,501 US11915410B2 (en) 2022-07-21 2023-04-24 Method and apparatus for inspecting tab appearance of cell assembly, and electronic device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/107073 WO2024016266A1 (zh) 2022-07-21 2022-07-21 检测电芯组件的极耳外观的方法与装置、电子设备

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/138,501 Continuation US11915410B2 (en) 2022-07-21 2023-04-24 Method and apparatus for inspecting tab appearance of cell assembly, and electronic device

Publications (1)

Publication Number Publication Date
WO2024016266A1 true WO2024016266A1 (zh) 2024-01-25

Family

ID=89576772

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/107073 WO2024016266A1 (zh) 2022-07-21 2022-07-21 检测电芯组件的极耳外观的方法与装置、电子设备

Country Status (4)

Country Link
US (1) US11915410B2 (zh)
EP (1) EP4339887A1 (zh)
CN (1) CN117751381A (zh)
WO (1) WO2024016266A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20160087600A (ko) * 2015-01-14 2016-07-22 한화테크윈 주식회사 불량 검사 장치 및 방법
CN107941805A (zh) * 2017-12-01 2018-04-20 无锡先导智能装备股份有限公司 电芯质量检测方法
CN114609153A (zh) * 2022-04-15 2022-06-10 荣旗工业科技(苏州)股份有限公司 一种电池极耳缺陷检测设备及方法
CN114764804A (zh) * 2022-06-16 2022-07-19 深圳新视智科技术有限公司 锂电池极片缺陷检测方法、装置、设备及存储介质

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004087238A (ja) * 2002-08-26 2004-03-18 Nissan Motor Co Ltd 積層型電池
DE102007063184B4 (de) * 2007-08-06 2011-01-05 Daimler Ag Einzelzelle für eine Batterie zur elektrischen Kontaktierung
US8138567B2 (en) * 2008-04-18 2012-03-20 Invisage Technologies, Inc. Materials, fabrication equipment, and methods for stable, sensitive photodetectors and image sensors made therefrom
KR101334121B1 (ko) * 2009-10-26 2013-11-29 에스케이이노베이션 주식회사 전지의 전극 검사장치 및 방법
JP6158789B2 (ja) * 2012-03-30 2017-07-05 三洋電機株式会社 積層式電池
KR101981010B1 (ko) * 2018-08-01 2019-08-28 정연길 이차 전지 제조 설비 및 이를 사용한 이차 전지 제조 방법
CN112557390B (zh) * 2019-09-10 2023-02-24 惠州旭鑫智能技术有限公司 一种动力电池裸电芯极耳错位缺陷单目视觉检测方法
CN112330623B (zh) * 2020-10-30 2022-05-27 蜂巢能源科技有限公司 电芯极组极片对齐度检测方法和检测装置
CN215142258U (zh) 2021-08-04 2021-12-14 傲普(上海)新能源有限公司 一种锂电池电芯筛选工装
CN115035050A (zh) * 2022-05-23 2022-09-09 广东人工智能与先进计算研究院 一种电池极片表面缺陷检测方法、装置、设备及产品
CN115829921B (zh) * 2022-09-16 2024-01-05 宁德时代新能源科技股份有限公司 检测电芯缺陷的方法、装置和计算机可读存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20160087600A (ko) * 2015-01-14 2016-07-22 한화테크윈 주식회사 불량 검사 장치 및 방법
CN107941805A (zh) * 2017-12-01 2018-04-20 无锡先导智能装备股份有限公司 电芯质量检测方法
CN114609153A (zh) * 2022-04-15 2022-06-10 荣旗工业科技(苏州)股份有限公司 一种电池极耳缺陷检测设备及方法
CN114764804A (zh) * 2022-06-16 2022-07-19 深圳新视智科技术有限公司 锂电池极片缺陷检测方法、装置、设备及存储介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4339887A4 *

Also Published As

Publication number Publication date
US20240029233A1 (en) 2024-01-25
US11915410B2 (en) 2024-02-27
EP4339887A4 (en) 2024-03-20
EP4339887A1 (en) 2024-03-20
CN117751381A (zh) 2024-03-22

Similar Documents

Publication Publication Date Title
CN115829913B (zh) 裸电芯外观检测方法、装置、计算机设备和存储介质
WO2024108847A1 (zh) 辊、点检方法及装置、卷材传送方法及装置、设备、介质
WO2023193213A1 (zh) 电池极片绝缘涂层缺陷的检测方法、装置和计算机设备
CN114764804B (zh) 锂电池极片缺陷检测方法、装置、设备及存储介质
US20240077432A1 (en) Cell detection method, apparatus, and system, computer device, and storage medium
CN102818998B (zh) 一种锂离子动力电池析锂的检测方法
CN111024008B (zh) 一种叠片型储能器件的错位尺寸的检测方法及检测装置
EP4358207A1 (en) Electrode sheet surface adhesive pasting detection method, device and apparatus, and electrode sheet adhesive pasting machine
WO2023070552A1 (zh) 一种电池卷芯卷绕覆盖检测方法、装置及设备
CN115205198A (zh) 一种极耳质量检测方法、系统、设备及存储介质
CN104091322A (zh) 叠片锂离子电池的检测方法
WO2024016266A1 (zh) 检测电芯组件的极耳外观的方法与装置、电子设备
CN112629442B (zh) 一种叠片电芯的探测方法、装置、电子设备和存储介质
WO2024093644A1 (zh) 电芯检测系统及电芯检测方法
CN113654493A (zh) 一种叠片软包锂电池质量检测方法及系统
CN106091999B (zh) 一种极片错位的检测方法和装置
CN117173100B (zh) 聚合物锂离子电池生产控制系统及其方法
CN116934732A (zh) 一种光伏组件的检测方法、装置及电子设备
WO2023202209A1 (zh) 对复合料带的阴极极片进行折角检测的方法、装置和系统
WO2023197117A1 (zh) 一种极耳的检测方法及装置
CN116739997A (zh) 一种基于x射线成像的锂电池在线成像分析方法
CN115797254A (zh) 极片缺陷检测方法、装置、计算机设备和存储介质
KR20220087027A (ko) 전지셀 검사 시스템 및 전지셀 검사 방법
US20240166460A1 (en) Roller, inspection method and apparatus, coil conveying method and apparatus, device and medium
WO2024036951A1 (zh) 深度图像的校正方法及装置、电池外壳组件焊接检测方法

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 202280006741.3

Country of ref document: CN

ENP Entry into the national phase

Ref document number: 2022888600

Country of ref document: EP

Effective date: 20230508