WO2023070365A1 - 电池极耳检测方法、装置及存储介质 - Google Patents

电池极耳检测方法、装置及存储介质 Download PDF

Info

Publication number
WO2023070365A1
WO2023070365A1 PCT/CN2021/126675 CN2021126675W WO2023070365A1 WO 2023070365 A1 WO2023070365 A1 WO 2023070365A1 CN 2021126675 W CN2021126675 W CN 2021126675W WO 2023070365 A1 WO2023070365 A1 WO 2023070365A1
Authority
WO
WIPO (PCT)
Prior art keywords
tab
connected domain
sub
layers
tabs
Prior art date
Application number
PCT/CN2021/126675
Other languages
English (en)
French (fr)
Inventor
陈璨
黄强威
王智玉
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/CN2021/126675 priority Critical patent/WO2023070365A1/zh
Priority to EP21958587.4A priority patent/EP4209778A4/en
Priority to CN202180066184.XA priority patent/CN116349021A/zh
Priority to US18/126,906 priority patent/US11823370B2/en
Publication of WO2023070365A1 publication Critical patent/WO2023070365A1/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
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/12—Edge-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M4/00—Electrodes
    • H01M4/02—Electrodes composed of, or comprising, active material
    • H01M4/04—Processes of manufacture in general
    • 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
    • G06T2207/30124—Fabrics; Textile; Paper
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02E60/10—Energy storage using batteries

Definitions

  • the present application relates to the technical field of battery assembly, in particular to a battery tab detection method, device and storage medium.
  • the manufacturing methods of power battery bare cells mainly include winding and stacking, among which winding is the most widely used method.
  • the battery pole pieces used for winding usually need to be die-cut, and only the part of the metal foil that needs to pass the current is kept, and this part of the metal foil is the tab. Due to the extremely thin thickness of the metal foil used in the lithium battery pole piece, its own strength is low, so the tabs are prone to be folded during the winding process, and thus be involved in the coating film area of the battery cell.
  • the folding situation is divided into two types: one is to fold into the pole piece where it is located; the other is to fold to the separator.
  • the present application provides a battery tab detection method, device and storage medium, which can automatically identify the folding of tabs, not only improve the detection efficiency, but also effectively reduce the missed detection rate.
  • the present application provides a battery tab detection method, including: obtaining a cross-sectional view of a multi-layer tab of the battery to be tested; identifying and analyzing the cross-sectional view to obtain a plurality of connected domains, each of the connected domains Including one tab or a plurality of tabs adhered to each other; according to the position and number of intersection points of the tabs in each connected domain, determine the number of tab layers corresponding to each connected domain; according to each of the connected domains The number of tab layers corresponding to the connected domain, calculate the total number of tab layers of the multi-layer tab in the cross-sectional view; judge the multi-layer tab according to the total number of tab layers and the preset real number of tab layers Whether there is a fold in the tab of the layer.
  • each connected domain includes a tab or Multiple interconnected tabs; according to the position and number of intersection points of tabs in each connected domain, each connected domain is analyzed separately, corresponding to the number of tabs in the connected domain, so that the obtained The number of tab layers is more accurate; and the total number of tab layers in the multi-layer tab in the cross-sectional view is determined according to the obtained number of tab layers corresponding to each connected domain.
  • the technical solution provided by this application first proposes a technical solution that uses connected domains to detect and determine the intersection point of tab adhesion, which utilizes the possible existence of tab adhesion
  • the cross-sectional view of the obtained multi-layer tab is divided into multiple connected domains, and then the tabs in each connected domain are calculated separately, and then the calculated results are summed up, so that the obtained cross-sectional view of the pole
  • the result of total ear layers is more accurate. Therefore, when judging whether there is a folded multi-layer tab according to the total number of layers of the tab and the preset real number of layers of the tab, the judgment accuracy is high, which can effectively reduce the missed detection rate.
  • the battery tab detection method in this embodiment can automatically identify the folded tab by introducing connected domains and performing related image recognition, which not only improves the detection efficiency, but also effectively reduces the missed detection rate.
  • determining the number of tab layers corresponding to each of the connected domains according to the position and number of intersection points of tab adhesion in each of the connected domains includes: identifying each of the connected domains The position and the number of the intersecting points of the tabs in the domain; if the number of intersections in the connected domain is 0, then determine that the number of tabs in the connected domain is 1; if the number of the intersecting points in the connected domain is 1 or more, the connected domain is divided into multiple regions according to the position of the intersection point, and each of the regions does not include the intersection point, and the connected domain is a plurality of interconnected tabs; according to The number of sub-connected domains in each of the plurality of regions determines the number of tab layers of the connected domain, and each of the sub-connected domains includes non-adhesive parts of the plurality of interconnected tabs or fully adhered parts.
  • the said connected domain is divided into a plurality of areas according to the position of the intersection, and each of the areas does not include the intersection, including: two positions of each intersection Taking a preset offset along the extending direction of the connected domain as a region boundary; dividing the connected domain into the plurality of regions according to the region boundary, and each of the regions does not include the intersection point.
  • the preset offset is taken as the area boundary along the extension direction of the connected domain on both sides of the position of each intersection point, which avoids dividing the ear part near the intersection point into the area, and avoids due to each area division. If it is too narrow, the tabs at the boundary will be counted repeatedly, thereby further improving the detection accuracy of the number of sub-connected domains in the region.
  • the identifying and analyzing the cross-sectional view obtains a plurality of connected domains, each of which includes a tab or a plurality of interconnected tabs, including: eliminating the area in the cross-sectional view smaller than Preset the tab adhesion holes with a threshold value to obtain the first image; perform connected domain analysis on the first image to obtain the plurality of connected domains, each of which includes one tab or a plurality of interconnected tabs .
  • eliminating tab adhesion holes whose area in the cross-sectional view is smaller than a preset threshold can improve efficiency.
  • the eliminating tab adhesion holes with an area smaller than a preset threshold in the cross-sectional view to obtain the first image includes: performing binarization processing on the cross-sectional view to obtain a binarized image; Setting the pixels at the position of the connected domain whose area is smaller than the preset threshold in the binarized image to white to obtain the first image.
  • the setting the pixels in the connected domain positions whose area is smaller than the preset threshold in the binarized image to white to obtain the first image includes: converting the binarized image Interchange the pixel values of pixels with a pixel value of 0 and pixels with a pixel value of 255 to obtain an exchanged image; determine the position of the connected domain with a pixel value of 255 and an area smaller than the preset threshold in the exchanged image; The first image is obtained by adjusting the pixel values of the pixels in the binarized image corresponding to the positions of the connected domains in the swapped image whose area is smaller than the preset threshold value to 255.
  • the determining the number of ear layers of the connected domain according to the number of sub-connected domains in each of the multiple areas includes: determining each of the multiple areas The number of inner sub-connected domains; taking the maximum number of sub-connected domains in the plurality of regions in the connected domain as the number of ear layers of the connected domain.
  • the maximum value of the number of sub-connected domains in multiple regions in the connected domain is taken as the number of tab layers in the connected domain, which is conducive to accurately calculating the number of layers of tabs in the area, thereby accurately detecting whether the tabs are A flip occurs.
  • the determining the number of sub-connected domains in each of the multiple areas includes: obtaining the lengths of all sub-connected domains in each of the multiple areas; according to each The number of sub-connected domains whose length in the region is greater than 1/N times of the longest sub-connected domain in the region determines the number of sub-connected domains in the region, and the N is greater than 0.
  • This embodiment proposes a method of determining the number of sub-connected domains in each region according to the number of sub-connected domains whose length is greater than 1/N times of the longest sub-connected domain in the region, so as to avoid glitches and Misjudged as the tab of adhesion, thereby eliminating the burr interference, and further improving the detection accuracy of the number of tab layers.
  • the present application provides a battery tab detection device, including: at least one processor; and a memory connected in communication with the at least one processor; wherein, the memory stores information that can be used by the at least one Instructions executed by a processor, the instructions are executed by the at least one processor, so that the at least one processor can execute the method for detecting battery tabs in the foregoing embodiments.
  • the present application provides a computer-readable storage medium storing a computer program, which is characterized in that, when the computer program is executed by a processor, the battery tab detection method as in the above-mentioned embodiment is implemented.
  • Fig. 1 is a schematic diagram of several adhesion states found by the inventor
  • Fig. 2 is a schematic structural diagram of a battery tab detection system in some embodiments of the present application.
  • FIG. 3 is a schematic flow chart of a battery tab detection method in some embodiments of the present application.
  • Fig. 4 is a cross-sectional view of a battery tab in some embodiments of the present application.
  • Fig. 5 is a schematic diagram after dividing the cross-sectional view shown in Fig. 4 into connected domains;
  • Fig. 6 is a schematic diagram after region division of the cross-sectional view shown in Fig. 4;
  • Fig. 7 is a schematic flow chart of eliminating tab adhesion holes in a battery tab detection method in some embodiments of the present application.
  • Fig. 8 is a sectional view in the flow chart shown in Fig. 7 of the present application.
  • Fig. 9 is the binarized image obtained after binarizing the sectional view in the flow chart shown in Fig. 7 of the present application.
  • Fig. 10 is a schematic diagram of burrs appearing on battery tabs in some embodiments of the present application.
  • Fig. 11 is a schematic flow chart of an example of a battery tab detection method in some embodiments of the present application.
  • FIG. 12 is a schematic structural diagram of a battery tab detection device in some embodiments of the present application.
  • multiple refers to more than two (including two), similarly, “multiple groups” refers to more than two groups (including two), and “multiple pieces” refers to More than two pieces (including two pieces).
  • the battery tabs Due to the fact that the battery tabs are folded into their own pole pieces, it will cause low capacity, short circuit and other phenomena in the battery cell, and what's more, it will cause thermal runaway and fire. Therefore, in the production process, it is necessary to detect whether the battery tab is turned over. In the current detection method, manual visual inspection is inefficient and prone to missed detection.
  • the Hi-pot test only has a high detection rate for the folded tabs on the isolation film, but it is difficult for the tabs folded into its own layer. identify. At present, there is no detection method that can automatically identify the folding of the tabs, which can not only improve the detection efficiency, but also effectively reduce the missed detection rate.
  • the tabs have the following adhesion states, as shown in Figure 1 (a solid line in Figure 1 represents one tab or multiple tabs that are completely adhered),
  • a single tab 111 is not adhered to other tabs at all, and there is no intersection point on the tab 111; in case 2, two tabs 112 and 113 and more than two tabs (not shown) automatically A certain intersection point 110 is not separated after adhesion, and there is one intersection point 110 on the two tabs 112 and 113; in the third case, two tabs 114 and 115 and two tabs (not shown) from a certain After the intersection points are adhered, the position of the self-intersection point is separated again, and there is one intersection point 120 on the two tabs 114 and 115; in case 4, two tabs 116 and 117 and more than two tabs (not shown) from the first After the first intersection point 130 is adhered, it is not separated temporarily, and it is separated from the
  • the applicant has given the following several embodiments, so that on the basis of realizing the automatic recognition of the folding of the tabs and improving the detection efficiency, the above-mentioned tab adhesions can be fully considered when determining the number of tab layers In the case of the battery tab, the number of tab layers of the battery tab can be accurately detected to reduce the missed detection rate.
  • the detection system includes: a processing device 1, an imaging device 2 and a transmission device, wherein the transmission device includes: a conveyor belt 3 and One or more trays 4 located on the conveyor belt 3, the trays 4 are used to place the batteries 5 to be tested.
  • the camera of the camera device 3 is aimed at the battery 5 to be tested on the tray 4 to take a cross-sectional view of the multi-layer tabs on the battery 5 to be tested.
  • the battery tab detection method disclosed in the embodiment of the present application can be used in the above-mentioned battery tab detection device or system, but is not limited to, and the above is only an example.
  • the battery tab detection method includes:
  • Step S11 Obtain a cross-sectional view of the multi-layer tab of the battery to be tested.
  • Step S12 Identify and analyze the cross-sectional view to obtain a plurality of connected domains, each connected domain includes one tab or multiple tabs adhered to each other.
  • Step S13 Determine the number of tab layers corresponding to each connected domain according to the position and number of intersection points of the tabs in each connected domain.
  • Step S14 According to the number of tab layers corresponding to each connected domain, calculate the total number of tab layers in the multi-layer tab in the cross-sectional view.
  • Step S15 According to the total number of layers of the tab and the preset real number of layers of the tab, it is judged whether there is a fold over the multi-layered tab.
  • connected Domain refers to the image area composed of foreground pixels with the same pixel value and adjacent positions in the image. Since there is no connected area between tabs that are not adhered to each other in the cross-sectional view, only a single tab or multiple tabs that are adhered to each other have a connected area. Therefore, by analyzing the connected domains of the cross-sectional view in Figure 4, multiple connected domains that are not adhered to each other in the multi-layer tab can be found, and one connected domain represents a tab, or multiple tabs that are adhered to each other.
  • Fig. 5 is a schematic diagram of dividing the cross-sectional view shown in Fig. 4 into connected domains, each dotted line box represents a connected domain, and identifying Fig. 4 can obtain 6 connected domains (1, 2, 3, 4, 5 and 6), According to the position and number of the intersection points of the tabs in each connected domain, the number of tab layers corresponding to the connected domain is determined, so that the obtained number of tab layers in each connected domain is more accurate; and according to each connected The number of tab layers corresponding to the domain is used to determine the total number of tab layers in the multi-layer tab in the cross-sectional view.
  • the technical solution provided by this application firstly proposes the use of connected
  • the technical scheme of detecting and judging the intersection points of the tab adhesions which utilizes the intersection points of the possible tab adhesions, divides the cross-sectional diagram of the obtained multi-layer tabs into multiple connected domains, and then analyzes each connected domain.
  • the lugs are calculated separately, and then the calculated results are added together, so that the result of the total number of layers of the lug in the cross-sectional view is more accurate.
  • the battery tab detection method in this embodiment can automatically identify the folded tab by introducing connected domains and performing related image recognition, which not only improves the detection efficiency, but also effectively reduces the missed detection rate.
  • determining the number of tab layers corresponding to each connected domain includes: identifying the intersection points of the tab adhesions in each connected domain Position and number; if the number of intersection points in the connected domain is 0, determine the number of ear layers in the connected domain as 1; if the number of intersection points in the connected domain is 1 or more, divide the connected domain into multiple regions according to the position of the intersection points , and each area does not include intersection points, the connected domain is a plurality of interconnected tabs; the number of tab layers of the connected domain is determined according to the number of sub-connected domains in each area in multiple areas, and each sub-connected domain includes multiple The non-adhesive part or the completely adhered part of the tabs that are adhered to each other.
  • the skeleton of the connected domain can be extracted.
  • the skeleton here can be understood as the lines of the connected domain, and the position and number of intersection points on the skeleton can be calculated. According to the position and number to determine.
  • the connected domain is a single tab, and the number of tab layers in the connected domain can be directly determined to be 0. As shown in Figure 5, if the number of intersection points in the first connected domain 1 is 0, then the first connected domain 1 includes a single tab 11; if the number of intersection points in the fourth connected domain 4 is 0, then the fourth connected domain 1 Field 4 includes a single tab 16 .
  • the connected domain is a plurality of interconnected tabs.
  • the intersection point 101 that exists in the second connected domain 2 the number of intersection points is 1, and this second connected domain 2 includes two mutually sticking tabs 12 and 13;
  • the intersection point that exists in the third connected domain 3 102, the number of intersections is 1, and the third connected domain 3 includes two interconnected tabs 14 and 15;
  • the connected domain 5 includes two tabs 17 and 18 that are glued to each other;
  • the number of intersections 106, 107 and 108 that exist in the sixth connected domain 6 is 3, and the sixth connected domain 6 includes four tabs that are glued to each other 19, 20, 21 and 22.
  • each sub-connected domain of each region includes multiple interconnected poles. Partially non-adhesive or fully adhered parts of the ear allow for accurate determination of the number of layers of the ear in each area.
  • FIG. 6 is a schematic diagram of dividing the cross-sectional view shown in FIG. 4 into regions, wherein each frame with a thin dashed line represents a connected domain, and each frame with a thick dashed line represents a region.
  • intersection 101 in the second connected domain 2 which can be divided into two regions 21 and 22 along the extension direction of the tab length, one of which includes two sub-connected domains 201 and 202, and each sub-connected domain is a tab Parts that are not adhered to each other, wherein the sub-connected domain 201 is the part where the tab 12 is not adhered to the tab 13, and the sub-connected domain 202 is the part where the tab 13 is not adhered to the tab 12; another area 22 includes a sub-connected domain 203 , the sub-connected domain 203 is a part where the tab 12 and the tab 13 are completely adhered.
  • the third connected domain 3 which can be divided into two regions 31 and 32 along the extension direction of the tab length, wherein one region 31 includes two sub-connected domains 301 and 302, and each sub-connected domain is a tab Parts that are not adhered to each other, wherein, the sub-connected domain 301 is the part where the tab 14 is not adhered to the tab 15, and the sub-connected domain 302 is the part where the tab 15 is not adhered to the tab 14; the other area 32 also includes two sub-regions.
  • each sub-connected domain is a part where the tabs are not adhered to each other, wherein, the sub-connected domain 303 is the part where the tab 14 is not adhered to the tab 15, and the sub-connected domain 304 is the part where the tab 15 is not connected to the tab 14 Adhesive part.
  • the fifth connected domain 5 can be divided into 4 regions 51, 52, 53 and 54 along the extending direction of the tab length, wherein, the two regions 51 and 53 include two sub-connected domains, and the region 51 includes sub-connected domains.
  • the sub-connected domain 501 is the part where the tab 17 is not adhered to the tab 18
  • the sub-connected domain 502 is the part where the tab 18 is not adhered to the tab 17
  • the area 53 includes sub-connected domains 504 and 505
  • the sub-connected domain 504 is the part where the tab 18 is not adhered to the tab 17, and the sub-connected domain 505 is the part where the tab 17 is not adhered to the tab 18
  • the other two areas 52 and 54 all include a sub-connected domain, wherein the area 52 includes a sub-connected domain 503, and the region 54 includes a sub-connected domain 506, and both the sub-connected domains 503 and 506 are parts of the complete adhesion of the tabs 17 and 18.
  • the sixth connected domain 6 which can be divided into four regions 61, 62, 63 and 64 along the extension direction of the tab length, and the regions 61 and 62 each include four sub-connected domains, wherein, The area 61 includes sub-connected domains 601, 602, 603 and 604, the area 62 includes sub-connected domains 605, 606, 607 and 608, and each sub-connected domain is a non-adhesive part of the tabs 19, 20, 21 and 22; Area 63 and area 64 both include three sub-connected domains, wherein, area 63 includes sub-connected domains 609, 610 and 611, the sub-connected domain 609 is the part where the tab 19 and the tab 20 are completely adhered, and the sub-connected domain 610 is the part of the tab.
  • the sub-connected domain 611 is the tab 22; the area 64 includes sub-connected domains 612, 613 and 614, the sub-connected domain 612 is the part where the tab 19 and the tab 20 are completely adhered, the sub-connected domain 613 is the tab 21, and the sub-connected domain 613 is the tab 21.
  • the connected domain 614 is the tab 22 .
  • the connected domain is divided into multiple areas according to the position of the intersection, and each area does not include the intersection, including: extending the connected domain along the two sides of the position of each intersection The direction takes the preset offset as the region boundary; the connected domain is divided into multiple regions according to the region boundary, and each region does not include intersection points.
  • FIG 6 it is a schematic diagram of dividing the cross-sectional view shown in Figure 4 by taking the area boundary.
  • the short solid line perpendicular to the length extension direction of the connected domain indicates the area boundary.
  • a certain offset is taken as the area boundary according to the extension direction of the tab (that is, the extension direction of the connected domain) on both sides of each intersection point, and the two ends of the connected domain are automatically determined as the area boundary.
  • the preset offset can be set according to actual needs, for example, the preset offset can be 5 pixels, 8 pixels, 10 pixels, etc.
  • the preset offset is taken as the area boundary along the extension direction of the connected domain on both sides of the position of each intersection point, which avoids dividing the ear part near the intersection point into the area, and avoids due to each area division. If it is too narrow, the tabs at the boundary will be counted repeatedly, thereby further improving the detection accuracy of the number of sub-connected domains in the region.
  • each connected domain includes one tab or multiple interconnected tabs, including: eliminating the area in the cross-sectional view smaller than the preset Thresholded tab adhesion holes to obtain the first image; conducting connected domain analysis on the first image to obtain a plurality of connected domains, each connected domain includes one tab or multiple tabs adhered to each other.
  • the preset threshold can be set according to actual needs, for example, the preset threshold can be 25 pixels, 20 pixels, 25 pixels, etc.
  • Step S21 Binarize the cross-sectional view to obtain a binary image, wherein the cross-sectional view is shown in FIG. 4 , and the binarized image is shown in FIG. 8 .
  • Step S22 Swap the pixel values of pixels with a pixel value of 0 and pixels with a pixel value of 255 in the binarized image to obtain a swapped image, as shown in FIG. 9 .
  • Step S23 Determine the position of the connected domain in the swapped image with a pixel value of 255 and an area smaller than a preset threshold.
  • Step S24 Adjust the pixel values of the pixels in the binarized image to 255 corresponding to the connected domain positions in the swapped image whose area is smaller than the preset threshold to obtain the first image.
  • eliminating tab adhesion holes with an area smaller than a preset threshold in the cross-sectional view to obtain the first image includes: performing binarization processing on the cross-sectional view to obtain a binarized image;
  • the first image is obtained by setting the pixels at the position of the connected domain whose area is smaller than the preset threshold in the binarized image to white.
  • the cross-sectional view can be binarized to obtain a binary image.
  • the pixel value of the pixel in the binarized image is 255, which is displayed as white ;
  • the pixel value of the pixel in the background part is 0, which is displayed as black.
  • the black connected domain whose area is smaller than the preset threshold can be confirmed as the polar ear adhesion hole.
  • the pixel value becomes 255, so that the color of the tab adhesion hole is the same as that of the tab, and the tab adhesion hole and the tab are visually integrated to eliminate the tab adhesion hole.
  • setting the pixels at the positions of the connected domain whose area is smaller than the preset threshold in the binarized image to white to obtain the first image includes: setting pixels with a pixel value of 0 in the binarized image and The pixel value of the pixel whose pixel value is 255 is exchanged to obtain the exchanged image; the position of the connected domain whose pixel value is 255 in the exchanged image and whose area is smaller than the preset threshold is determined; the area in the corresponding exchanged image in the binarized image is less than The pixel value of the pixel at the position of the connected domain of the preset threshold is adjusted to 255 to obtain the first image.
  • the pixels at the positions of the black connected domains whose area is smaller than the preset threshold can be directly set to white to eliminate the tab adhesion holes.
  • Another implementation is given in this embodiment.
  • the pixel values of the pixels with a pixel value of 0 and the pixels with a pixel value of 255 in the binarized image can be exchanged first to obtain an exchanged image, and the polar ear in the exchanged image Some pixels with a pixel value of 0 are displayed as black, and some background pixels with a pixel value of 255 are displayed as white.
  • the first image is obtained by identifying white connected domains whose area is smaller than a preset threshold, and adjusting the pixel value of the white connected domain whose area corresponding to the swapped image is smaller than the preset threshold in the binarized image to 255.
  • determining the number of ear layers of the connected domain according to the number of sub-connected domains in each of the multiple areas includes: determining the number of sub-connected domains in each of the multiple areas; The maximum value of the number of sub-connected domains in multiple regions in the connected domain is taken as the number of ear layers of the connected domain.
  • the second connected domain 2 in accompanying drawing 5 is divided into two areas 21 and 22, and the area 21 includes two sub-connected domains 201 and 202, and the area 22 includes For one sub-connected domain 203, the maximum number of sub-connected domains in the two regions 21 and 22 is 2, then the number of tab layers of the second connected domain 2 is 2 layers.
  • the third connected domain 3 is divided into 2 areas 31 and 32, including 2 sub-connected domains 301 and 302 in the area 31, including 2 sub-connected domains 303 and 304 in the area 32, and the sub-connected domains in these 2 areas 31 and 32
  • the maximum value of the number of is 2, then the number of tab layers of the second connected domain 2 is 2 layers.
  • the fifth connected domain 5 is divided into four areas 51, 52, 53 and 54, the area 51 includes 2 sub-connected domains 501 and 502, the area 52 includes 1 sub-connected domain 503, and the area 53 includes 2 sub-connected domains 504 and 505, one sub-connected domain 506 is included in the region 54, the maximum number of the sub-connected domains in the four regions 51, 52, 53 and 54 is 2, then the number of ear layers of the fifth connected domain 5 is 2 layer.
  • the sixth connected domain 6 is divided into 4 areas 61, 62, 63 and 64, the area 61 includes 4 sub-connected domains 601, 602, 603 and 604, and the area 62 includes 4 sub-connected domains 605, 606, 607 and 608 , the area 63 includes 3 sub-connected domains 609, 610 and 611, and the area 64 includes 3 sub-connected domains 612, 6130 and 614, the maximum number of sub-connected domains in these 4 areas 61, 62, 63 and 64 is 4, then the number of tab layers of the sixth connected domain 6 is 4 layers.
  • the maximum value of the number of sub-connected domains is taken as the number of ear layers of the connected domain.
  • the maximum value of the number of sub-connected domains in multiple regions in the connected domain is taken as the number of tab layers in the connected domain, which is conducive to accurately calculating the number of layers of tabs in the area, thereby accurately detecting whether the tabs are A flip occurs.
  • determining the number of sub-connected domains in each area in the multiple areas includes: obtaining the lengths of all sub-connected domains in each area in the multiple areas; according to the length in each area The number of sub-connected domains greater than 1/N times of the longest sub-connected domain in the region determines the number of sub-connected domains in the region, and N is greater than 0.
  • step size how many pixels are counted once in the vertical direction
  • the processing method is: after dividing each connected domain into regions, determine the length of the sub-connected domains in each region, determine the maximum length of all sub-connected regions in this region, and when sub-connected regions appear in this region When the length of the domain is less than the maximum value of 1/N*length, the sub-connected domain is not counted within the number of sub-connected domains in this area. For example, suppose that area 1001 in Figure 10 includes two sub-connected domains 10 and 20, The length of the connected domain 10 is L2, and the length of the sub-connected domain 20 is L1.
  • the sub-connected domain 20 is not counted; only the length in the area 1001 is greater than or A sub-connected domain 10 equal to 1/N*L2, wherein N is greater than 0, N is an empirical value, and N usually takes a value between 5 and 6.
  • the number of sub-connected domains in the region is determined to avoid misjudgment of burrs as sticky ears. Therefore, the burr interference is eliminated, and the detection accuracy of the number of tab layers is further improved.
  • each connected domain includes a tab or multiple tabs that are connected to each other.
  • Binarization processing is to change the gray value of the point on the image to 0 or 255, that is, to make the entire image appear obvious black and white effect.
  • the pixel value of the pixel point in the ear part in the cross-sectional view is changed to 255, which is displayed as white; the pixel value of the pixel point in the background part is changed to 0, which is displayed as black.
  • each connected domain includes at least one tab or multiple tabs adhered to each other.
  • each connected domain may be a tab or a background interference, which can be screened according to the length characteristics of the tab, for example: the length of the tab is generally a fixed value, and the connected domain with the length of the fixed value is roughly the same It can be determined as the required connected domain, and the connected domain whose length is too short or too long compared with the fixed value can be determined as interference noise.
  • S33 Take a connected domain, and determine the position and number of intersection points in the connected domain.
  • the position and number of intersection points in the connected domain can be calculated by extracting the skeleton of the connected domain.
  • step S6 determines the number of layers of multiple interconnected tabs in the connected domain.
  • S36 Divide the connected domain into multiple areas according to the positions of the intersection points, and each area does not include the intersection points.
  • the two tabs with the intersection part will be regarded as a sub-connected domain in the subsequent judgment of the number of sub-connected domains in the area, which will lead to sub-connected domains in the area Number detection is not accurate. Therefore, it is necessary to divide the connected domain into multiple regions along the extension direction of the tab length according to the position of the intersection point, and each region does not include the position of the intersection point, which is conducive to accurately determining the number of layers of the tab in each region.
  • the preset offset is taken as the area boundary along the extension direction of the connected domain on both sides of the position of each intersection point, which avoids dividing the ear part near the intersection point into the area, so as to further improve the sub-connected domain in the area. Number of detection accuracy.
  • N is greater than 0, N is an empirical value, and N usually takes a value between 5 and 6.
  • the number of tab layers in some areas in the cross-sectional view increases, while the number of tab layers in some areas decreases, because in order to accurately detect whether the tabs are turned over, it is necessary to take the data in multiple areas in the connected domain.
  • the maximum value of the number of sub-connected domains is taken as the number of ear layers of the connected domain.
  • step S38 Determine whether all connected domains have been traversed. If the determination is yes, then execute step S9; otherwise, return to step S3 for execution;
  • step S8 After executing steps S5 and S7, execute step S8 to judge whether all connected domains have been traversed, and if all connected domains have been traversed, continue to execute step S9 to further determine the total number of layers of multi-layered tabs in the cross-sectional view. If not all the connected domains have been traversed, step S3 needs to be continued to analyze the connected domains for which the number of tab layers has not been determined.
  • the total number of tab layers in the multi-layer tab in the cross-sectional view can be obtained by adding the number of tab layers corresponding to each connected domain.
  • the number of tab layers corresponding to the first connected domain is 1, the number of tab layers corresponding to the second connected domain is 2, the number of tab layers corresponding to the third connected domain is 2, and the number of tab layers corresponding to the fourth connected domain is 2.
  • the number of tab layers corresponding to the connected domain is 1, the number of tab layers corresponding to the fifth connected domain is 2, and the number of tab layers corresponding to the sixth connected domain is 4, then the cross-sectional view corresponding to Figure 5 is multi-layered in Figure 4
  • the total number of layers of the tabs is 12.
  • the number of layers of the battery tab When the battery tab is turned over, the number of layers of the battery tab will increase. By detecting the number of layers of the battery tab and judging whether the number of layers of the battery tab is the same as the actual number of layers of the actual tab, it is determined that the battery Whether the tabs are turned over.
  • the total number of layers of the multi-layer tab in the cross-sectional view is 12. Assuming that the real number of layers of the tab is 12, the total number of layers of the tab is the same as the preset real number of layers of the tab. There is no turning over of the tab; assuming that the real number of layers of the tab is 10, there is a turning over of the multi-layer tab in the cross-sectional view.
  • the battery lug detection method in this embodiment aiming at the phenomenon of possible adhesion of the tabs, analyzes the pattern of intersection points generated by the adhesion, and proposes to divide the connected domain into multiple regions by taking the connected domain as the unit and the intersection points of the tab adhesion , and conduct connected domain analysis for each region, and count the number of tabs, compared with directly identifying the total number of tab layers in the cross-sectional view, the result of the total number of tab layers in the cross-sectional view is more accurate; therefore, when When judging whether there is a folded multi-layer tab according to the total number of layers of the tab and the preset real number of layers of the tab, the judgment accuracy is high, which can effectively reduce the missed detection rate.
  • the method for detecting battery tabs in this embodiment can automatically identify the folding of the tabs, which not only improves the detection efficiency, but also effectively reduces the missed detection rate.
  • the detection device includes: at least one processor 701; and a memory 702 communicatively connected to the at least one processor 701; Wherein, the memory 702 stores instructions that can be executed by at least one processor 701, and the instructions are executed by at least one processor 701, so that at least one processor 701 can execute the battery tab detection method in all embodiments of the present application.
  • the device shown in Figure 3 does not include a camera, therefore, an external camera can be used to take pictures, and the battery tab detection device directly obtains the photographed multi-layer tabs of the battery to be tested from the external camera. Sectional view.
  • the memory 702 and the processor 701 are connected by a bus, and the bus may include any number of interconnected buses and bridges, and the bus connects one or more processors 701 and various circuits of the memory 702 together.
  • the bus may also connect together various other circuits such as peripherals, voltage regulators, and power management circuits, all of which are well known in the art and therefore will not be further described herein.
  • the bus interface provides an interface between the bus and the transceivers.
  • a transceiver may be a single element or multiple elements, such as multiple receivers and transmitters, providing means for communicating with various other devices over a transmission medium.
  • the data processed by the processor 701 is transmitted on the wireless medium through the antenna, and further, the antenna also receives the data and transmits the data to the processor 701 .
  • the processor 701 is responsible for managing the bus and general processing, and may also provide various functions including timing, peripheral interface, voltage regulation, power management, and other control functions. And the memory 702 may be used to store data used by the processor 701 when performing operations.
  • the computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the battery tab detection method in any of the above-mentioned embodiments is implemented.
  • a storage medium includes several instructions to make a device ( It may be a single-chip microcomputer, a chip, etc.) or a processor (processor) to execute all or part of the steps of the methods described in the various embodiments of the present application.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disc, etc., which can store program codes. .

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Theoretical Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Chemical & Material Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • General Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Electrochemistry (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Manufacturing & Machinery (AREA)
  • Connection Of Batteries Or Terminals (AREA)

Abstract

本申请公开了一种电池极耳检测方法，包括：获取待检测电池的多层极耳的截面图；识别并分析所述截面图得到多个连通域，每个所述连通域包括一个极耳或多个相互粘连的极耳；根据每个所述连通域中极耳粘连的交点的位置及数目，确定每个所述连通域对应的极耳层数；根据每个所述连通域对应的极耳层数，计算所述截面图中所述多层极耳的极耳总层数；根据所述极耳总层数与预设的极耳真实层数，判断所述多层极耳是否存在翻折。本申请提供的电池极耳检测方法、装置及存储介质，能够自动识别极耳翻折情况，不仅提高了检测效率，还能够有效降低漏检率。

Description

电池极耳检测方法、装置及存储介质 技术领域
本申请涉及电池装配技术领域，具体涉及一种电池极耳检测方法、装置及存储介质。
背景技术
随着电动车的逐渐增多，动力电池的应用也日益增加。目前，动力电池裸电芯的制造方法主要包括卷绕和叠片两种，其中，应用最广泛的为卷绕方式。为了便于生产制造，卷绕所使用的电芯极片通常需要模切处理，仅保留需要通过电流的部分金属箔材，该部分金属箔材即为极耳。由于锂电池极片使用的金属箔材厚度极薄，自身强度较低，因此在卷绕过程中极耳容易发生翻折，从而被卷入电芯涂膜区。而翻折的情况分为两种类型：一种是翻折进自身所在极片中；另一种是翻折到隔离膜上。现有的极耳翻折一般通过目检以及施加电压对电芯阻值监测的Hi-pot测试进行检测。但是人工目检效率低下且容易发生漏检，Hi-pot测试只对翻折到隔离膜上的极耳翻折检出率高，而对翻折进自身层内的极耳难以辨识。而这种极耳翻折进自身极片的情况，会导致电芯产生低容、短路等现象，更有甚者，会引起热失控着火。
发明内容
鉴于上述问题，本申请提供一种电池极耳检测方法、装置及存储介质，能够自动识别极耳翻折情况，不仅提高了检测效率，还能够有效降低漏检率。
第一方面，本申请提供了一种电池极耳检测方法，包括：获取待检测电池的多层极耳的截面图；识别并分析所述截面图得到多个连通域，每个所述连通域包括一个极耳或多个相互粘连的极耳；根据每个所述连通域中极耳粘连的交点的位置及数目，确定每个所述连通域对应的极耳层数；根据每个所述连通域对应的极耳层数，计算所述截面图中所述多层极耳的极耳总层数；根据所述极耳总层数与预设的极耳真实层数，判断所述多层极耳是否存在翻折。
本申请实施例的技术方案中的电池极耳检测方法，在获取待检测电池的多层极耳的截面图后，识别并分析截面图得到多个连通域，每个连通域包括一个极耳或多个相互粘连的极耳；根据每个连通域中极耳粘连的交点的位置及数目对每个连通域单独进行分析对应该连通域的极耳层数，使得得到的每个连通域中的极耳层数更加准确；且根据得到的每个连通域对应的极耳层数来确定截面图中多层极耳的极耳总层数。相比于直接识别截面图中极耳总层数来说，本申请提供的技术方案首先提出了利用连通域进行检测和判定极耳粘连的交点的技术方案，其利用了可能存在的极耳粘连的交点，将获得的多层极耳的截面图分为了多个连通域，之后对每个连通域内的极耳进行单独的计算，再将计算的结果加和，从而使得得到的截面图中极耳总层数的结果更加准确。因此当根据极耳 总层数与预设的极耳真实层数判断多层极耳是否存在翻折时，判断准确率较高，能够有效降低漏检率。另外，本实施例中的电池极耳检测方法通过引入连通域以及进行相关的图像识别能够自动识别极耳翻折情况，不仅提高了检测效率，还能够有效降低漏检率。
在一些实施例中，所述根据每个所述连通域中极耳粘连的交点的位置及数目，确定每个所述连通域对应的极耳层数，包括：识别所述每个所述连通域中极耳粘连的交点的位置及数目；若所述连通域中所述交点数目为0，则确定所述连通域的极耳层数为1；若所述连通域中所述交点数目为1个及以上，则根据所述交点的位置将所述连通域划分为多个区域，且每个所述区域均不包括所述交点，所述连通域为多个相互粘连的极耳；根据所述多个区域中每个所述区域内子连通域的数目确定所述连通域的极耳层数，每个所述子连通域包括所述多个相互粘连的极耳中互不粘连的部分或完全粘连的部分。本实施例中，当连通域中存在交点，则说明连通域中存在多个相互粘连的极耳，此时需根据交点的位置沿极耳长度延伸方向将连通域划分为多个区域，由于每个区域均不再包括交点，因此，得到的多个区域中不会由于某个区域中存在极耳部分粘连，而将两个粘连在一起的部分视为一个子连通域的情况，有利于准确判定每个区域内极耳的层数。
在一些实施例中，所述根据所述交点的位置将所述连通域划分为多个区域，且每个所述区域均不包括所述交点，包括：对每一个所述交点的位置的两侧沿所述连通域的延伸方向取预设偏移量作为区域边界；根据所述区域边界将所述连通域划分为所述多个区域，且每个所述区域均不包括所述交点。本实施例中对每一个交点的位置的两侧沿连通域的延伸方向取预设偏移量作为区域边界，避免了将交点附近的极耳部分划分进区域内，以避免由于每个区域划分得太窄造成边界处的极耳被重复计数，从而进一步提高区域内子连通域数目的检测准确性。
在一些实施例中，所述识别并分析所述截面图得到多个连通域，每个所述连通域包括一个极耳或多个相互粘连的极耳，包括：消除所述截面图中面积小于预设阈值的极耳粘连孔以得到第一图像；对所述第一图像进行连通域分析得到所述多个连通域，每个所述连通域包括一个极耳或多个相互粘连的极耳。本实施例中消除截面图中面积小于预设阈值的极耳粘连孔，可提高效率。
在一些实施例中，所述消除所述截面图中面积小于预设阈值的极耳粘连孔以得到第一图像，包括：对所述截面图进行二值化处理得到二值化图像；将所述二值化图像中面积小于所述预设阈值的连通域位置处的像素置为白色以得到所述第一图像。
在一些实施例中，所述将所述二值化图像中面积小于所述预设阈值的连通域位置处的像素置为白色以得到所述第一图像，包括：将所述二值化图像中像素值为0的像素和像素值为255的像素的像素值互换得到互换图像；确定所述互换图像中像素值为255、且面积小于所述预设阈值的连通域位置；将所述二值化图像中对应所述互换图像中面积小于所述预设阈值的连通域位置处的像素的像素值调整为255以得到所述第一图像。
在一些实施例中，所述根据所述多个区域中每个所述区域内子连通域的数目确定所述连通 域的极耳层数，包括：确定所述多个区域中每个所述区域内子连通域的数目；取所述连通域中所述多个区域中子连通域的数目的最大值作为所述连通域的极耳层数。本实施例中取连通域中多个区域中子连通域的数目的最大值作为该连通域的极耳层数，有利于准确计算出区域内极耳的层数，从而准确检测出极耳是否发生翻折。
在一些实施例中，所述确定所述多个区域中每个所述区域内子连通域的数目，包括：获取所述多个区域中每个所述区域内所有子连通域的长度；根据每个所述区域内长度大于所述区域内长度最长的子连通域的1/N倍的子连通域的数目，确定所述区域内子连通域的数目，所述N大于0。本实施例中提出通过根据每个区域内长度大于区域内长度最长的子连通域的1/N倍的子连通域的数目，确定区域内子连通域的数目的方法，实现了避免将毛刺也误判为粘连的极耳，从而消除了毛刺干扰，进一步提高了极耳层数的检测准确性。
第二方面，本申请提供了一种电池极耳检测装置，包括：至少一个处理器；以及，与所述至少一个处理器通信连接的存储器；其中，所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行如上述实施例中的电池极耳检测方法。
第三方面，本申请提供了一种计算机可读存储介质，存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如上述实施例中的电池极耳检测方法。
上述说明仅是本申请技术方案的概述，为了能够更清楚了解本申请的技术手段，而可依照说明书的内容予以实施，并且为了让本申请的上述和其它目的、特征和优点能够更明显易懂，以下特举本申请的具体实施方式。
附图说明
通过阅读对下文优选实施方式的详细描述，各种其他的优点和益处对于本领域普通技术人员将变得清楚明了。附图仅用于示出优选实施方式的目的，而并不认为是对本申请的限制。而且在全部附图中，用相同的附图标号表示相同的部件。在附图中：
图1为发明人发现的极耳存在的几种粘连状态的示意图；
图2为本申请一些实施例中电池极耳检测系统的结构示意图；
图3为本申请一些实施例中电池极耳检测方法的流程示意图；
图4为本申请一些实施例中的电池极耳截面图；
图5为将图4所示截面图进行连通域划分后的示意图；
图6为对图4所示的截面图进行区域划分后的示意图；
图7为本申请一些实施例中电池极耳检测方法消除极耳粘连孔的流程示意图；
图8为本申请图7所示流程图中的截面图；
图9为本申请图7所示流程图中将截面图二值化后得到的二值化图像；
图10为本申请一些实施例中电池极耳出现毛刺的示意图；
图11为本申请一些实施例中电池极耳检测方法的示例流程示意图；
图12为本申请一些实施例中电池极耳检测装置的结构示意图。
具体实施方式
下面将结合附图对本申请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申请的技术方案，因此只作为示例，而不能以此来限制本申请的保护范围。
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申请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，不是旨在于限制本申请；本申请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。
在本申请实施例的描述中，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。在本申请实施例的描述中，“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申请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。
在本申请实施例的描述中，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。
在本申请实施例的描述中，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)。
在本申请实施例的描述中，技术术语“中心”“纵向”“横向”“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申请实施例和简化描述，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申请实施例的限制。
在本申请实施例的描述中，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申请实施例中的具体含义。
由于电池极耳翻折进自身极片的情况，会导致电芯产生低容、短路等现象，更有甚者，会引起热失控着火。因此，在生产过程中需要对电池极耳是否出现翻折的情况进行检测。目前的检测方法中人工目检效率低下且容易发生漏检，Hi-pot测试只对翻折到隔离膜上的极耳翻折检出率高，而对翻折进自身层内的极耳难以辨识。目前还没有一种能够自动识别极耳翻折情况，不仅能够提高检测效率，还能够有效降低漏检率的检测方法。
为了缓解上述问题，申请人研究发现，对于型号固定的电池来说，其电池极耳的层数是固定不变的，而当电池极耳出现翻折时，电池极耳的层数会增多，例如：当一个极耳翻折时，其层数由原来的1层变为2层。因此，可通过检测电池极耳的层数，并判断电池极耳的层数与实际的真实极耳层数是否相同来确定该电池中的极耳是否出现翻折。
基于此，申请人发现通过拍摄待检测电池的多层极耳的截面图像，分析截面图像的像素亮度和分布，能够得到待检测电池的多层极耳对应的若干线条图，可随机设置多个采样点，获取每个采样点上的线条信息，取其中的最大值作为多层极耳的统计结果。然而，申请人经过实践发现：在实际生产过程中，由于极耳箔片会出现粘连，而当采样点落在粘连位置时，会影响对截面分割后极耳层数的统计，导致统计的极耳层数不准确。
基于以上考虑，为了解决问题，发明人经过深入研究发现：极耳存在以下几种粘连状态，如图1所示(图1中一条实线表示一个极耳或多个完全粘连的极耳)，第①种情况，单个极耳111完全没有与其他极耳粘连，其极耳111上不存在交点；第②种情况，2片极耳112和113及2片以上极耳(未示出)自某一交点110粘连后没有再分开，2片极耳112和113上存在1个交点110；第③种情况，2片极耳114和115及以2片上极耳(未示出)自某一交点粘连后，自交点位置再次分开，2片极耳114和115上存在1个交点120；第④种情况，2片极耳116和117及2片以上极耳(未示出)自第一个交点130粘连后，暂时未分开，自第二个交点140分开，如果在极耳中间位置40处再次粘连(图中所示情况)，则2片极耳116和117上总共会存在3个交点130、140和150；如果在极耳根部(靠近极片一侧)或尾部分开(未示出)则其线条图上总共会存在2个交点；第⑤种情况，组合模式，前面几种模式的极耳之间出现不同位置的粘连，粘连交接点的数目取决于粘连产生的位置，图1中4片极耳118、119、120和121上总共存在3个交点160、170和180。
基于此，申请人给出了以下几种实施例，使得在实现自动识别极耳翻折情况，提高检测效率的基础上，同时还能够在确定极耳层数时充分考虑上述几种极耳粘连的情况，准确检测电池极耳 的极耳层数，降低漏检率。
首先，对于实现本申请所有实施例检测方法的装置和系统进行说明。
参见图2，对于实现本申请所提供的实施例检测方法可选的一种检测系统进行说明，该检测系统包括：处理装置1、摄像装置2以及传送装置，其中，传送装置包括：传送带3和位于传送带3上的一个或多个托盘4，托盘4用于放置待检测电池5。摄像装置3的摄像头对准其中托盘4上的待检测电池5，以拍摄待检测电池5上多层极耳的截面图，处理装置1与摄像装置2连接，用于获取摄像装置2拍摄到的待检测电池5的多层极耳的截面图，并实现本申请实施例中电池极耳检测方法以自动识别极耳翻折情况。本申请实施例公开的电池极耳检测方法可以但不限用于上述电池极耳检测装置或系统中，上述仅为举例说明。
其次，对本申请的电池极耳检测方法的实施例进行说明。
根据本申请的一些实施例的电池极耳检测方法，如图3所示，包括：
步骤S11：获取待检测电池的多层极耳的截面图。
步骤S12：识别并分析截面图得到多个连通域，每个连通域包括一个极耳或多个相互粘连的极耳。
步骤S13：根据每个连通域中极耳粘连的交点的位置及数目，确定每个连通域对应的极耳层数。
步骤S14：根据每个连通域对应的极耳层数，计算截面图中多层极耳的极耳总层数。
步骤S15：根据极耳总层数与预设的极耳真实层数，判断多层极耳是否存在翻折。
假设获取到的待检测电池的多层极耳的截面图如图4所示，其中一条实线表示一个极耳或多个完全粘连的极耳，识别并分析截面图得到多个连通域，连通域是指图像中具有相同像素值且位置相邻的前景像素点组成的图像区域。由于截面图中互不粘连的极耳之间不存在连通的区域，只有单个极耳或多个相互粘连的极耳才存在连通的区域。因此，通过对截面图图4进行连通域分析便可找将多层极耳中互不粘连的多个连通域，一个连通域代表一个极耳，或者多个相互粘连的极耳。
图5为将图4所示截面图进行连通域划分后的示意图，每一个虚线框表示一个连通域，识别图4可得到6个连通域(1、2、3、4、5和6)，根据每个连通域中极耳粘连的交点的位置及数目来确定该连通域对应的极耳层数，使得得到的每个连通域中的极耳层数更加准确；且根据得到的每个连通域对应的极耳层数来确定截面图中多层极耳的极耳总层数，相比于直接识别截面图中极耳总层数来说，本申请提供的技术方案首先提出了利用连通域进行检测和判定极耳粘连的交点的技术方案，其利用了可能存在的极耳粘连的交点，将获得的多层极耳的截面图分为了多个连通域，之后 对每个连通域内的极耳进行单独的计算，再将计算的结果加和，从而使得得到的截面图中极耳总层数的结果更加准确。因此当根据极耳总层数与预设的极耳真实层数判断多层极耳是否存在翻折时，判断准确率较高，能够有效降低漏检率。另外，本实施例中的电池极耳检测方法通过引入连通域以及进行相关的图像识别能够自动识别极耳翻折情况，不仅提高了检测效率，还能够有效降低漏检率。
根据本申请的一些实施例，根据每个连通域中极耳粘连的交点的位置及数目，确定每个连通域对应的极耳层数，包括：识别每个连通域中极耳粘连的交点的位置及数目；若连通域中交点数目为0，则确定连通域的极耳层数为1；若连通域中交点数目为1个及以上，则根据交点的位置将连通域划分为多个区域，且每个区域均不包括交点，连通域为多个相互粘连的极耳；根据多个区域中每个区域内子连通域的数目确定连通域的极耳层数，每个子连通域包括多个相互粘连的极耳中互不粘连的部分或完全粘连的部分。
在确定每个连通域中对应的极耳层数时，可提取该连通域的骨架，这里的骨架可理解为连通域的线条，并计算骨架上的交点的位置和数目，根据交点的位置和数目来确定。
若一个连通域中的交点数目为0个，则可以确定该连通域为单个极耳，可直接确定该连通域中的极耳层数为0。如图5所示，第一连通域1中的交点数目为0个，则该第一连通域1包括单个极耳11；第四连通域4中的交点数目为0个，则该第四连通域4包括单个极耳16。
若一个连通域中的交点数目为1个及以上，则该连通域为多个相互粘连的极耳。如图5所示，第二连通域2中存在的交点101，交点数目为1个，该第二连通域2包括两个相互粘连的极耳12和13；第三连通域3中存在的交点102，交点数目为1个，该第三连通域3包括两个相互粘连的极耳14和15；第五连通域5中存在的交点103、104和105，交点数目为3个，该第五连通域5包括两个相互粘连的极耳17和18；第六连通域6中存在的交点106、107和108，交点数目为3个，该第六连通域6包括四个相互粘连的极耳19、20、21和22。
若连通域为多个相互粘连的极耳，需根据交点的位置沿极耳长度的延伸方向将连通域划分为多个区域，由于每个区域均不包括交点位置，因此，得到的多个区域中不会由于某个区域中存在极耳部分粘连，而将两个粘连在一起的部分视为一个子连通域的情况，本实施例每个区域的每个子连通域包括多个相互粘连的极耳中互不粘连的部分或完全粘连的部分，有利于准确判定每个区域内极耳的层数。
如图6所示为将图4所示的截面图进行区域划分后的示意图，其中，每一个细虚线框代表一个连通域，每个一加粗虚线框代表一个区域。
第二连通域2内存在1个交点101，沿极耳长度延伸方向可划分为2个区域21和22，其中一个区域21包括两个子连通域201和202，且每个子连通域均为极耳互不粘连的部分，其中，子连通域201为极耳12不与极耳13粘连的部分，子连通域202为极耳13不与极耳12粘连的部分；另一个区域22包括一个子连通域203，该子连通域203为极耳12与极耳13完全粘连的部分。
第三连通域3内存在1个交点102，沿极耳长度延伸方向可划分为2个区域31和32，其中一个区域31包括两个子连通域301和302，且每个子连通域均为极耳互不粘连的部分，其中，子连通域301为极耳14不与极耳15粘连的部分，子连通域302为极耳15不与极耳14粘连的部分；另一个区域32也包括两个子连通域，且每个子连通域均为极耳互不粘连的部分，其中，子连通域303为极耳14不与极耳15粘连的部分，子连通域304为极耳15不与极耳14粘连的部分。
第五连通域5内存在3个交点，沿极耳长度延伸方向可划分为4个区域51、52、53和54，其中，两个区域51和53均包括两个子连通域，区域51包括子连通域501和502，子连通域501为极耳17不与极耳18粘连的部分，子连通域502为极耳18不与极耳17粘连的部分；区域53包括子连通域504和505，子连通域504为极耳18不与极耳17粘连的部分，子连通域505为极耳17不与极耳18粘连的部分；另两个区域52和54均包括一个子连通域，其中区域52包括子连通域503，区域54包括子连通域506，子连通域503和506均为极耳17和18完全粘连的部分。
第六连通域6内存在3个交点106、107和108，沿极耳长度延伸方向可划分为4个区域61、62、63和64，区域61和区域62均包括四个子连通域，其中，区域61包括子连通域601、602、603和604，区域62包括子连通域605、606、607和608，且每个子连通域均为极耳19、20、21和22互不粘连的部分；区域63和区域64均包括三个子连通域，其中，区域63包括子连通域609、610和611，子连通域609为极耳19和极耳20完全粘连的部分，子连通域610为极耳21，子连通域611为极耳22；区域64包括子连通域612、613和614，子连通域612为极耳19和极耳20完全粘连的部分，子连通域613为极耳21，子连通域614为极耳22。
根据本申请的一些实施例，可选地，根据交点的位置将连通域划分为多个区域，且每个区域均不包括交点，包括：对每一个交点的位置的两侧沿连通域的延伸方向取预设偏移量作为区域边界；根据区域边界将连通域划分为多个区域，且每个区域均不包括交点。
如图6所示为将图4所示的截面图取区域边界进行区域划分后的示意图，图6中与连通域的长度延伸方向垂直的短实线表示区域边界，在根据交点位置将连通域划分为多个区域时，对每一个交点两侧按极耳的延展方向(即连通域的延伸方向)取一定的偏移量作为区域边界，连通域的两 端自动确定为区域边界，根据区域边界将连通域划分为多个均不包括交点的区域。其中，预设偏移量可根据实际需求自行设置，例如：预设偏移量可为5个像素点、8个像素点、10个像素点等。
由于交点附近的极耳部分可能依然存在粘连，若直接以交点所在位置进行区域划分，则确定出的区域内的子连通域数目可能依然不准确。本实施例中对每一个交点的位置的两侧沿连通域的延伸方向取预设偏移量作为区域边界，避免了将交点附近的极耳部分划分进区域内，以避免由于每个区域划分得太窄造成边界处的极耳被重复计数，从而进一步提高区域内子连通域数目的检测准确性。
根据本申请的一些实施例，可选地，识别并分析截面图得到多个连通域，每个连通域包括一个极耳或多个相互粘连的极耳，包括：消除截面图中面积小于预设阈值的极耳粘连孔以得到第一图像；对第一图像进行连通域分析得到多个连通域，每个连通域包括一个极耳或多个相互粘连的极耳。
如图6所示的第五连通域5中，当出现极耳粘连孔(图6中第五连通域5的区域53所示位置处)时，若该极耳粘连孔的面积足够小，则可消除该极耳粘连孔，以提高检测的效率。且消除该小面积的极耳粘连孔后，由于极耳在靠近电池的一端(图6中左侧)必然不会粘连，因此，不会影响该连通域极耳层数的确定。其中，该预设阈值可以根据实际需求自行设置，例如：预设阈值可为25个像素点、20个像素点、25个像素点等。
上述消除截面图中面积小于预设阈值的极耳粘连孔以得到第一图像的具体过程如图7所示，包括：
步骤S21：对截面图进行二值化处理得到二值化图像，其中，截面图如图4所示，二值化图像如图8所示。
步骤S22：将二值化图像中像素值为0的像素和像素值为255的像素的像素值互换得到互换图像，互换图像如图9所示。
步骤S23：确定互换图像中像素值为255、且面积小于预设阈值的连通域位置。
步骤S24：将二值化图像中对应互换图像中面积小于预设阈值的连通域位置处的像素的像素值调整为255以得到第一图像。
根据本申请的一些实施例，消除截面图中面积小于预设阈值的极耳粘连孔以得到第一图像，包括：对截面图进行二值化处理得到二值化图像；
将二值化图像中面积小于预设阈值的连通域位置处的像素置为白色以得到第一图像。
在消除截面图中可能出现的极耳粘连孔时，可先对截面图进行二值化处理得到二值化图 像，该二值化图像中极耳部分像素点的像素值为255，显示为白色；背景部分像素点的像素值为0，显示为黑色。二值化图像中面积小于预设阈值的黑色连通域即可确认为极耳粘连孔，通过将该面积小于预设阈值的黑色连通域所在位置处的像素点置为白色，即像素值变为255，使得极耳粘连孔的颜色与极耳的颜色相同，极耳粘连孔与极耳在视觉上融为一体，以消除极耳粘连孔。
根据本申请的一些实施例，将二值化图像中面积小于预设阈值的连通域位置处的像素置为白色以得到第一图像，包括：将二值化图像中像素值为0的像素和像素值为255的像素的像素值互换得到互换图像；确定互换图像中像素值为255、且面积小于预设阈值的连通域位置；将二值化图像中对应互换图像中面积小于预设阈值的连通域位置处的像素的像素值调整为255以得到第一图像。
在上述消除极耳粘连孔时，可直接将面积小于预设阈值的黑色连通域所在位置处的像素点置为白色以消除极耳粘连孔。本实施例中给出了另一种实现方式，可先将二值化图像中像素值为0的像素和像素值为255的像素的像素值互换得到互换图像，互换图像中极耳部分像素点的像素值为0显示为黑色，背景部分像素点的像素值为255显示为白色。通过识别面积小于预设阈值的白色连通域，并将二值化图像中对应互换图像的面积小于预设阈值的白色连通域所在位置处的像素值调整为255以得到第一图像。
根据本申请的一些实施例，可选地，根据多个区域中每个区域内子连通域的数目确定连通域的极耳层数，包括：确定多个区域中每个区域内子连通域的数目；取连通域中多个区域中子连通域的数目的最大值作为连通域的极耳层数。
结合附图4、附图5和附图6进行说明，附图5中第二连通域2被划分为2个区域21和22，区域21中包括2个子连通域201和202，区域22中包括一个子连通域203，这2个区域21和22中子连通域的数目的最大值为2，则该第二连通域2的极耳层数为2层。
第三连通域3被划分为2个区域31和32，区域31中包括2个子连通域301和302，区域32中包括2个子连通域303和304，这2个区域31和32中子连通域的数目的最大值为2，则该第二连通域2的极耳层数为2层。
第五连通域5被划分为4个区域51、52、53和54，区域51中包括2个子连通域501和502，区域52中包括1个子连通域503，区域53中包括2个子连通域504和505，区域54中包括1个子连通域506，这4个区域51、52、53和54中子连通域的数目的最大值为2，则该第五连通域5的极耳层数为2层。
第六连通域6被划分为4个区域61、62、63和64，区域61中包括4个子连通域601、 602、603和604，区域62中包括4个子连通域605、606、607和608，区域63中包括3个子连通域609、610和611，区域64中包括3个子连通域612、6130和614，这4个区域61、62、63和64中子连通域的数目的最大值为4，则该第六连通域6的极耳层数为4层。
由于极耳发生翻折时，截面图中部分区域极耳层数增多，而部分区域极耳层数减小，因此为准确检测出极耳是否发生翻折，需取连通域中多个区域中子连通域的数目的最大值作为该连通域的极耳层数。本实施例中取连通域中多个区域中子连通域的数目的最大值作为该连通域的极耳层数，有利于准确计算出区域内极耳的层数，从而准确检测出极耳是否发生翻折。
根据本申请的一些实施例，可选地，确定多个区域中每个区域内子连通域的数目，包括：获取多个区域中每个区域内所有子连通域的长度；根据每个区域内长度大于区域内长度最长的子连通域的1/N倍的子连通域的数目，确定区域内子连通域的数目，所述N大于0。
如图10所示，若某连通域1中极耳11表面出现毛刺200，则在根据交点100划分的区域1001和1002中，区域1001存在毛刺200容易确定出的子连通域的数目大于该区域1001真实的极耳层数，导致最终的检测结果不准确。
发明人发现若通过在待检测电池多层极耳的线条图上设置采样点来确定极耳层数时，会产生采样点间隔(步长)如何选择的问题，即隔多少像素统计一次垂直方向极耳的数目，若步长选择偏大，则极耳局部分开未被统计，容易导致计数少于真实值，误判为极耳翻折；步长小一定程度上可以降低漏统计的问题，但对局部异常分割产生的毛刺会更敏感，容易多计数，其次，增加了程序的计算量。
本实施例中在统计每个连通域的极耳层数之前，需要消除局部的毛刺，防止将毛刺误当极耳统计。处理方法为：在将每个连通域进行区域划分后，对每个区域中的子连通域的长度进行确定，确定出该区域内所有子连通区域的长度最大值，当该区域内出现子连通域的长度小于1/N*长度最大值时，在统计在该区域内的子连通域数目内，不统计该子连通域，例如假设图10中区域1001包括两个子连通域10和20，子连通域10的长度为L2，子连通域20的长度为L1，若子连通域20的长度L1小于1/N*L2时，则不统计该子连通域20；只统计该区域1001内长度大于或等于1/N*L2的子连通域10，其中，N大于0，N为经验值，N通常取5～6之间的数值。
根据每个区域内长度大于区域内长度最长的子连通域的1/N倍的子连通域的数目，确定区域内子连通域的数目，实现了避免将毛刺也误判为粘连的极耳，从而消除了毛刺干扰，进一步提高了极耳层数的检测准确性。
下面是本申请的一些实施例，参见图11，具体包括：
S31:获取待检测电池的多层极耳的截面图。
S32:识别并分析截面图得到多个连通域，每个连通域包括一个极耳或多个相互粘连的极耳。
在得到截面图后，对截面图进行二值化处理得到二值化图像，二值化处理即是将图像上的点的灰度值变为0或255，也就是将整个图像呈现出明显的黑白效果。本实施例中将截面图中的极耳部分像素点的像素值变为255，显示为白色；背景部分像素点的像素值变为0，显示为黑色。
对二值化图像进行连通域分析得到多个连通域，每个连通域包括至少一个极耳或多个相互粘连的极耳。其中，每一个连通域可能是极耳也可能是背景干扰，可根据极耳的长度特征进行筛选，例如：极耳的长度一般为一个固定的值，与该固定的值长度大致相同的连通域即可确定为所需的连通域，与该固定的值相比长度过短或过长的连通域可确定为干扰噪声。
S33:取一个连通域，并确定连通域中交点的位置和数目。
可通过提取该连通域的骨架以计算该连通域中交点的位置和数目。
S34：判断连通域中交点的数目是否大于0，若不大于0，则执行步骤S5；否则，执行步骤S6。
S35：确定该连通域对应的极耳层数为1。
参见图5，若一个连通域中的交点数目为0个，则可以确定该连通域为单个极耳，可直接确定该连通域中的极耳层数为0。若一个连通域中的交点数目为1个及以上，则该连通域为多个相互粘连的极耳，此时，需继续执行步骤S6进一步判断连通域中多个相互粘连的极耳的层数。
S36:根据交点的位置将连通域划分为多个区域，且每个区域均不包括交点。
若某个区域中存在两个极耳相交的交点部分，在后续判断区域内的子连通域数目时会将该存在交点部分的两个极耳视为一个子连通域，会导致区域内子连通域数目检测不准确性。因此，需根据交点的位置沿极耳长度延伸方向将连通域划分为多个区域，每个区域均不包括交点位置，从而有利于准确判定每个区域内极耳的层数。
可选地，由于交点附近的极耳部分可能依然存在粘连，若直接以交点所在位置进行区域划分，则确定出的区域内的子连通域数目可能依然不准确。本实施例中对每一个交点的位置的两侧沿连通域的延伸方向取预设偏移量作为区域边界，避免了将交点附近的极耳部分划分进区域内，以进一步提高区域内子连通域数目的检测准确性。
S37：根据每个区域内长度大于区域内长度最长的子连通域的1/N倍的子连通域的数目，确定区域内子连通域的数目。
由于若某连通域中极耳表面出现毛刺，则在根据交点划分的区域中，存在毛刺的区域中容易确定出的子连通域的数目大于该区域真实的极耳层数，导致最终的检测结果不准确。因此，本实施例中在统计每个连通域的极耳层数之前，需要消除局部的毛刺，防止将毛刺误当极耳统计。具体处理方式如下：
对每个区域中的子连通域的长度进行确定，确定出该区域内所有子连通区域的长度最大值max_length，当该区域内出现子连通域的长度小于
1/N*max_length时，在统计在该区域内的子连通域数目内，不统计该子连通域；只统计该区域内长度大于或等于1/N*max_length的子连通域。其中，N大于0，N为经验值，N通常取5～6之间的数值。
由于极耳发生翻折时，截面图中部分区域极耳层数增多，而部分区域极耳层数减小，因为为准确检测出极耳是否发生翻折，需取连通域中多个区域中子连通域的数目的最大值作为该连通域的极耳层数。
S38：判断是否遍历完所有的连通域。若判定为是，则执行步骤S9；否则，返回步骤S3执行；
在执行完步骤S5及S7之后，执行步骤S8判断是否遍历完所有的连通域，若遍历完所有的连通域，则继续执行步骤S9进一步确定截面图中多层极耳的极耳总层数。若未遍历完所有的连通域，则需继续执行步骤S3，继续对未确定极耳层数的连通域进行分析。
S39：根据每个连通域对应的极耳层数，计算截面图中多层极耳的极耳总层数。
将每个连通域对应的极耳层数相加便，可得到截面图中多层极耳的极耳总层数。以附图5所示为例，第一连通域对应的极耳层数为1，第二连通域对应的极耳层数为2，第三连通域对应的极耳层数为2，第四连通域对应的极耳层数为1，第五连通域对应的极耳层数为2，第六连通域对应的极耳层数为4，则图5所对应的截面图图4中多层极耳的极耳总层数为12。
S40：根据极耳总层数与预设的极耳真实层数，判断多层极耳是否存在翻折。
当电池极耳出现翻折时，电池极耳的层数会增多，通过检测电池极耳的层数，并判断电池极耳的层数与实际的真实极耳层数是否相同来确定该电池中的极耳是否出现翻折。上述例子中截面图中多层极耳的极耳总层数为12，假设极耳真实层数为12，则极耳总层数与预设的极耳真实层数相同，截面图中多层极耳不存在翻折；假设极耳真实层数为10，则截面图中多层极耳存在翻折。
本实施例的电池极耳检测方法，针对极耳可能出现粘连的现象，分析其粘连产生相交点的模式，提出了以连通域为单位，以极耳粘连的交点将连通域划分为多个区域，并对每个区域进行连 通域分析，统计极耳的数目，相比于直接识别截面图中极耳总层数来说，得到的截面图中极耳总层数的结果更加准确；因此当根据极耳总层数与预设的极耳真实层数判断多层极耳是否存在翻折时，判断准确率较高，能够有效降低漏检率。本实施例中的电池极耳检测方法能够自动识别极耳翻折情况，不仅提高了检测效率，还能够有效降低漏检率。另外，对极耳截面图中出现的一些异常毛刺，提出通过根据每个区域内长度大于区域内长度最长的子连通域的1/N倍的子连通域的数目，确定区域内子连通域的数目的方法，以消除毛刺干扰，提高了极耳层数的检测准确性，进一步降低漏检率。
经验证发现，针对产线生产电芯数据，利用本实施例的电池极耳检测方法判定极耳翻折缺陷过杀率在3％以内，漏杀率控制在0.7％以内；且实际产线数据产生这类毛刺的概率约在0.1％，利用本实施例的电池极耳检测方法在进入统计极耳层数的逻辑前，消除局部毛刺异常干扰，降低由于极耳分割异常出现毛刺而导致漏杀的情况。
参见图12，对于实现本申请所有实施例检测方法的一种电池极耳检测装置进行说明，该检测装置包括：包括至少一个处理器701；以及，与至少一个处理器701通信连接的存储器702；其中，存储器702存储有可被至少一个处理器701执行的指令，指令被至少一个处理器701执行，以使至少一个处理器701能够执行本申请所有实施例电池极耳的检测方法。图3所示的装置中并未包括摄像装置，因此，可利用外部的摄像装置进行拍摄，电池极耳检测装置直接从外部的摄像装置处获取其已拍摄的待检测电池的多层极耳的截面图。
其中，存储器702和处理器701采用总线方式连接，总线可以包括任意数目的互联的总线和桥，总线将一个或多个处理器701和存储器702的各种电路连接在一起。总线还可以将诸如外围设备、稳压器和功率管理电路等之类的各种其他电路连接在一起，这些都是本领域所公知的，因此，本文不再对其进行进一步描述。总线接口在总线和收发机之间提供接口。收发机可以是一个元件，也可以是多个元件，比如多个接收器和发送器，提供用于在传输介质上与各种其他装置通信的单元。经处理器701处理的数据通过天线在无线介质上进行传输，进一步，天线还接收数据并将数据传送给处理器701。
处理器701负责管理总线和通常的处理，还可以提供各种功能，包括定时、外围接口、电压调节、电源管理以及其他控制功能。而存储器702可以被用于存储处理器701在执行操作时所使用的数据。
根据本申请的一些实施例的计算机可读存储介质，存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如上述任一实施例中的电池极耳检测方法。
即，本领域技术人员可以理解，实现上述实施例方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成，该程序存储在一个存储介质中，包括若干指令用以使得一个设备(可以是单片机，芯片等)或处理器(processor)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是：以上各实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述各实施例对本申请进行了详细的说明，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申请各实施例技术方案的范围，其均应涵盖在本申请的权利要求和说明书的范围当中。尤其是，只要不存在冲突，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。本申请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 (10)

  1. 一种电池极耳检测方法，其特征在于，包括：
    获取待检测电池的多层极耳的截面图；
    识别并分析所述截面图得到多个连通域，每个所述连通域包括一个极耳或多个相互粘连的极耳；
    根据每个所述连通域中极耳粘连的交点的位置及数目，确定每个所述连通域对应的极耳层数；
    根据每个所述连通域对应的极耳层数，计算所述截面图中所述多层极耳的极耳总层数；
    根据所述极耳总层数与预设的极耳真实层数，判断所述多层极耳是否存在翻折。
  2. 如权利要求1所述的电池极耳检测方法，其特征在于，所述根据每个所述连通域中极耳粘连的交点的位置及数目，确定每个所述连通域对应的极耳层数，包括：
    识别所述每个所述连通域中极耳粘连的交点的位置及数目；
    若所述连通域中所述交点数目为0，则确定所述连通域的极耳层数为1；
    若所述连通域中所述交点数目为1个及以上，则根据所述交点的位置将所述连通域划分为多个区域，且每个所述区域均不包括所述交点，所述连通域为多个相互粘连的极耳；
    根据所述多个区域中每个所述区域内子连通域的数目确定所述连通域的极耳层数，每个所述子连通域包括所述多个相互粘连的极耳中互不粘连的部分或完全粘连的部分。
  3. 如权利要求2中所述的电池极耳检测方法，其特征在于，所述根据所述交点的位置将所述连通域划分为多个区域，且每个所述区域均不包括所述交点，包括：
    对每一个所述交点的位置的两侧沿所述连通域的延伸方向取预设偏移量作为区域边界；
    根据所述区域边界将所述连通域划分为所述多个区域，且每个所述区域均不包括所述交点。
  4. 如权利要求1所述的电池极耳检测方法，其特征在于，所述识别并分析所述截面图得到多个连通域，每个所述连通域包括一个极耳或多个相互粘连的极耳，包括：
    消除所述截面图中面积小于预设阈值的极耳粘连孔以得到第一图像；
    对所述第一图像进行连通域分析得到所述多个连通域，每个所述连通域包括一个极耳或多个相互粘连的极耳。
  5. 如权利要求4所述的电池极耳检测方法，其特征在于，所述消除所述截面图中面积小于预设阈值的极耳粘连孔以得到第一图像，包括：
    对所述截面图进行二值化处理得到二值化图像；
    将所述二值化图像中面积小于所述预设阈值的连通域位置处的像素置为白色以得到所述第一图像。
  6. 如权利要求5所述的电池极耳检测方法，其特征在于，所述将所述二值化图像中面积小于所述预设阈值的连通域位置处的像素置为白色以得到所述第一图像，包括：
    将所述二值化图像中像素值为0的像素和像素值为255的像素的像素值互换得到互换图像；
    确定所述互换图像中像素值为255、且面积小于所述预设阈值的连通域位置；
    将所述二值化图像中对应所述互换图像中面积小于所述预设阈值的连通域位置处的像素的像素值调整为255以得到所述第一图像。
  7. 如权利要求2所述的电池极耳检测方法，其特征在于，所述根据所述多个区域中每个所述区域内子连通域的数目确定所述连通域的极耳层数，包括：
    确定所述多个区域中每个所述区域内子连通域的数目；
    取所述连通域中所述多个区域中子连通域的数目的最大值作为所述连通域的极耳层数。
  8. 如权利要求7所述的电池极耳检测方法，其特征在于，所述确定所述多个区域中每个所述区域内子连通域的数目，包括：
    获取所述多个区域中每个所述区域内所有子连通域的长度；
    根据每个所述区域内长度大于所述区域内长度最长的子连通域的1/N倍的子连通域的数目，确定所述区域内子连通域的数目，所述N大于0。
  9. 一种电池极耳检测装置，其特征在于，包括：至少一个处理器；以及，
    与所述至少一个处理器通信连接的存储器；其中，
    所述存储器存储有可被所述至少一个处理器执行的指令，所述指令被所述至少一个处理器执行，以使所述至少一个处理器能够执行如权利要求1至8中任一所述的电池极耳检测方法。
  10. 一种计算机可读存储介质，存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如权利要求1至8中任一所述的电池极耳检测方法。
PCT/CN2021/126675 2021-10-27 2021-10-27 电池极耳检测方法、装置及存储介质 WO2023070365A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
PCT/CN2021/126675 WO2023070365A1 (zh) 2021-10-27 2021-10-27 电池极耳检测方法、装置及存储介质
EP21958587.4A EP4209778A4 (en) 2021-10-27 2021-10-27 METHOD AND APPARATUS FOR DETECTING BATTERY TAB, AND STORAGE MEDIUM
CN202180066184.XA CN116349021A (zh) 2021-10-27 2021-10-27 电池极耳检测方法、装置及存储介质
US18/126,906 US11823370B2 (en) 2021-10-27 2023-03-27 Method and apparatus for inspecting battery tab and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/126675 WO2023070365A1 (zh) 2021-10-27 2021-10-27 电池极耳检测方法、装置及存储介质

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US18/126,906 Continuation US11823370B2 (en) 2021-10-27 2023-03-27 Method and apparatus for inspecting battery tab and storage medium

Publications (1)

Publication Number Publication Date
WO2023070365A1 true WO2023070365A1 (zh) 2023-05-04

Family

ID=86160326

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/126675 WO2023070365A1 (zh) 2021-10-27 2021-10-27 电池极耳检测方法、装置及存储介质

Country Status (4)

Country Link
US (1) US11823370B2 (zh)
EP (1) EP4209778A4 (zh)
CN (1) CN116349021A (zh)
WO (1) WO2023070365A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116577345A (zh) * 2023-07-14 2023-08-11 广州市易鸿智能装备有限公司 一种锂电池极耳数量检测方法及系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117350993A (zh) * 2023-11-02 2024-01-05 上海贝特威自动化科技有限公司 一种基于图像识别的极耳层数检测方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108375544A (zh) * 2018-01-11 2018-08-07 多氟多（焦作）新能源科技有限公司 一种用于检测叠片式电池极耳弯折的方法
CN112525917A (zh) * 2020-11-12 2021-03-19 欣旺达电动汽车电池有限公司 电芯极耳检测方法、装置、系统及存储介质
CN113156528A (zh) * 2021-03-09 2021-07-23 欣旺达电动汽车电池有限公司 极耳翻折检测装置和检测方法
CN113203745A (zh) * 2021-07-05 2021-08-03 中航锂电科技有限公司 一种叠片装置及极片翻折的检测方法
CN113376177A (zh) * 2021-06-21 2021-09-10 上海商汤科技开发有限公司 极耳检测方法、装置及电子设备

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110530338A (zh) * 2019-09-18 2019-12-03 东莞塔菲尔新能源科技有限公司 一种极耳发生折叠的检测方法及装置
DE102020003892B4 (de) * 2020-06-29 2021-12-09 Daimler Ag Zellhalter für wenigstens eine Batteriezelle sowie Zellmodul
WO2022029923A1 (ja) * 2020-08-05 2022-02-10 オリンパス株式会社 医療機器用の物品容器の部品、医療機器用の物品容器、医療機器収納パック、医療機器用の物品
CN114204220B (zh) * 2020-08-31 2023-04-07 比亚迪股份有限公司 电芯、电池以及电池包
EP4166935B1 (en) * 2021-08-26 2024-02-14 Contemporary Amperex Technology Co., Limited Apparatus and method for detecting tab folding, and image analyzer
JP7412394B2 (ja) * 2021-08-27 2024-01-12 プライムプラネットエナジー＆ソリューションズ株式会社 電池
KR20230111908A (ko) * 2022-01-19 2023-07-26 에스케이온 주식회사 이차 전지

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108375544A (zh) * 2018-01-11 2018-08-07 多氟多（焦作）新能源科技有限公司 一种用于检测叠片式电池极耳弯折的方法
CN112525917A (zh) * 2020-11-12 2021-03-19 欣旺达电动汽车电池有限公司 电芯极耳检测方法、装置、系统及存储介质
CN113156528A (zh) * 2021-03-09 2021-07-23 欣旺达电动汽车电池有限公司 极耳翻折检测装置和检测方法
CN113376177A (zh) * 2021-06-21 2021-09-10 上海商汤科技开发有限公司 极耳检测方法、装置及电子设备
CN113203745A (zh) * 2021-07-05 2021-08-03 中航锂电科技有限公司 一种叠片装置及极片翻折的检测方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4209778A4 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116577345A (zh) * 2023-07-14 2023-08-11 广州市易鸿智能装备有限公司 一种锂电池极耳数量检测方法及系统
CN116577345B (zh) * 2023-07-14 2023-09-29 广州市易鸿智能装备有限公司 一种锂电池极耳数量检测方法及系统

Also Published As

Publication number Publication date
EP4209778A4 (en) 2023-12-27
US20230289948A1 (en) 2023-09-14
EP4209778A1 (en) 2023-07-12
CN116349021A (zh) 2023-06-27
US11823370B2 (en) 2023-11-21

Similar Documents

Publication Publication Date Title
WO2023070365A1 (zh) 电池极耳检测方法、装置及存储介质
US11455805B2 (en) Method and apparatus for detecting parking space usage condition, electronic device, and storage medium
WO2020042741A1 (zh) 电池检测方法及装置
TWI729606B (zh) 用於一邊緣運算網路的負載平衡裝置及方法
CN110049320B (zh) 摄像头遮挡检测方法、装置、电子设备及存储介质
US20230417823A1 (en) Fault detection circuit, method AND power adapter
US20220405145A1 (en) Method, Apparatus, System and Electronic Device for Selecting Intelligent Analysis Algorithm
US11875531B1 (en) Detection method, electronic device, and computer-readable storage medium
JP6347155B2 (ja) 画像処理装置、画像判断方法、及びプログラム
CN109887137A (zh) 一种起落杆的控制方法及装置
CN107403169B (zh) 信号灯检测识别方法及装置
CN113989626B (zh) 一种基于目标检测模型的多类别垃圾场景区分方法
CN110390813B (zh) 基于车型鉴别的大数据处理系统
WO2022227548A1 (zh) 抛洒事件检测方法、装置、电子设备、存储介质及计算机程序产品
US20220019818A1 (en) Method and system for vehicle parking detection, and storage medium
CN108957397A (zh) 巡检机器人的定位方法及装置
US11969903B1 (en) Control method, electronic device and storage medium
CN110535779B (zh) 流量降级方法及装置
CN116563841B (zh) 配电网设备标识牌的检测方法、检测装置和电子设备
CN111986473A (zh) 基于车型鉴别的大数据处理方法
CN112434717B (zh) 一种模型训练方法及装置
CN109299652B (zh) 一种用于图像定位的模型训练方法、图像定位方法及装置
CN112235925A (zh) 目视助航设备状态预警方法及相关装置
CN111862665A (zh) 管理数据即时显示平台
CN108377501B (zh) 一种tvws的频谱冲突检测方法、装置及设备

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2021958587

Country of ref document: EP

Effective date: 20230404