CN-122023284-A - Self-adaptive coated pole piece size detection method

CN122023284ACN 122023284 ACN122023284 ACN 122023284ACN-122023284-A

Abstract

The invention discloses a self-adaptive coated pole piece size detection algorithm, and belongs to the technical field of visual detection in lithium battery manufacturing. The method aims to solve the problems that the traditional visual detection algorithm needs to manually and frequently adjust parameters due to product model replacement, so that the efficiency is low and errors are easy to introduce. The method comprises the steps of S1, preprocessing an acquired pole piece image, S2, analyzing connected domains of the preprocessed image, intelligently identifying pole piece models based on the number of the connected domains, S3, automatically loading a corresponding predefined detection scheme according to the identified pole piece models, S4, extracting edge points and fitting edge straight lines in an area designated by the detection scheme, and S5, calculating the process size of the pole piece according to the fitted edge straight lines. The invention realizes the full-automatic self-adaptive switching of the detection algorithm, greatly improves the detection efficiency and accuracy of the production line, and reduces the labor cost.

Inventors

ZHANG LI
PAN XIAOBAO

Assignees

合肥国轩高科动力能源有限公司

Dates

Publication Date: 20260512
Application Date: 20260112

Claims (8)

1. The self-adaptive coated pole piece size detection method is characterized by comprising the following steps of: s1, performing image preprocessing on an acquired original image of a coated pole piece to obtain an optimized binary image; S2, carrying out connected domain analysis on the optimized binary image, and determining the number of material areas of different pole piece types so as to identify the pole piece type; S3, automatically loading a predefined detection scheme corresponding to the model according to the identification result by the system, wherein the detection scheme defines at least one detection area of the size to be detected and algorithm parameters; S4, in each detection area defined by the detection scheme, performing edge point accurate extraction and straight line fitting to obtain a straight line equation representing the edge of the pole piece; and S5, calculating to obtain each process size of the coated pole piece based on the fitted edge linear equation.
2. The method according to claim 1, wherein step S1 comprises the steps of: s11, filtering the original image by adopting a median filtering algorithm to remove impulse noise and protect edge details; S12, performing binarization processing on the filtered image, and converting the image into a black-white binary image; And S13, carrying out morphological open operation on the binary image so as to eliminate isolated noise points and smooth the edge of the target area.
3. The method of claim 1, wherein in step S2, a two-pass scanning algorithm is adopted to carry out connected domain marking on the binary image, temporary labels are distributed for foreground pixels in the image in the first pass of scanning, equivalence relations among the labels are recorded, labels in the connected domain are unified according to the equivalence relations in the second pass of scanning, and finally the number of the connected domains is determined according to the number of unique labels, so that the pole piece model is judged.
4. The method according to claim 1, wherein the step S4 comprises: S41, positioning a specific detection area which needs to be subjected to size measurement in the image according to a detection scheme; S42, extracting stable edge points in each detection area by adopting a method of combining gradient projection and variance filtering; and S43, performing linear fitting on all the stable edge points extracted in the step S42 by using a least square method to obtain an edge linear equation of the detection region.
5. The method according to claim 4, wherein step S42 specifically comprises the steps of: S421, calculating gradient intensity projection of the image in the horizontal direction and the vertical direction in the detection area, and roughly positioning a row or a column where the edge is located according to the peak value of the projection curve; s422, traversing pixel points with gradient values exceeding a preset threshold as candidate edge points, and calculating pixel value variance of each candidate edge point in the neighborhood of the candidate edge point; s423, judging the candidate edge points with the variance smaller than the set threshold value as stable edge points, and eliminating the points with the variance larger than or equal to the set threshold value.
6. The method according to claim 1, wherein in step S5, the distance between two parallel edge lines is calculated using a point-to-line distance formula.
7. A computer readable storage medium, characterized in that a computer program is stored on the medium, which computer program, when run, performs the method according to any one of claims 1 to 6.
8. A computer system comprising a processor, a storage medium having a computer program stored thereon, the processor reading from the storage medium and running the computer program to perform the method of any one of claims 1 to 6.

Description

Self-adaptive coated pole piece size detection method Technical Field The invention relates to the field of manufacturing of lithium battery pole pieces, in particular to a self-adaptive size detection method of a coated pole piece. Background In the manufacturing process of the lithium battery, the pole piece is used as a key component of the battery, the dimensional accuracy of the pole piece directly influences the performance and the reliability of a final product, and if the production process is improper, the size of the pole piece is possibly abnormal, so that the battery often cannot meet the production standard. In order to ensure that the quality of the pole piece meets high standards, a visual detection technique is generally adopted to strictly monitor the size of the pole piece. With the frequent replacement of product models of the production line, the traditional visual detection algorithm needs to be manually adjusted to adapt to new specification requirements, and the method is complex in flow, low in efficiency and easy to introduce human errors. Disclosure of Invention In order to solve the existing problems, the invention provides a self-adaptive coated pole piece size detection method, which comprises the following specific scheme: an adaptive coated pole piece size detection method comprises the following steps: s1, performing image preprocessing on an acquired original image of a coated pole piece to obtain an optimized binary image; S2, carrying out connected domain analysis on the optimized binary image, and determining the number of material areas of different pole piece types so as to identify the pole piece type; S3, automatically loading a predefined detection scheme corresponding to the model according to the identification result by the system, wherein the detection scheme defines at least one detection area of the size to be detected and algorithm parameters; S4, in each detection area defined by the detection scheme, performing edge point accurate extraction and straight line fitting to obtain a straight line equation representing the edge of the pole piece; and S5, calculating to obtain each process size of the coated pole piece based on the fitted edge linear equation. Preferably, step S1 comprises the steps of: s11, filtering the original image by adopting a median filtering algorithm to remove impulse noise and protect edge details; S12, performing binarization processing on the filtered image, and converting the image into a black-white binary image; And S13, carrying out morphological open operation on the binary image so as to eliminate isolated noise points and smooth the edge of the target area. Preferably, in step S2, a two-pass scanning algorithm is adopted to perform connected domain marking on the binary image, the first-pass scanning allocates temporary labels for foreground pixels in the image and records equivalent relationships among the labels, the second-pass scanning unifies the labels in the connected domain according to the equivalent relationships, and finally, the number of the connected domains is determined according to the number of unique labels, so that the pole piece model is judged. Preferably, the step S4 includes: S41, positioning a specific detection area which needs to be subjected to size measurement in the image according to a detection scheme; S42, extracting stable edge points in each detection area by adopting a method of combining gradient projection and variance filtering; and S43, performing linear fitting on all the stable edge points extracted in the step S42 by using a least square method to obtain an edge linear equation of the detection region. Preferably, the step S42 specifically includes the steps of: S421, calculating gradient intensity projection of the image in the horizontal direction and the vertical direction in the detection area, and roughly positioning a row or a column where the edge is located according to the peak value of the projection curve; s422, traversing pixel points with gradient values exceeding a preset threshold as candidate edge points, and calculating pixel value variance of each candidate edge point in the neighborhood of the candidate edge point; s423, judging the candidate edge points with the variance smaller than the set threshold value as stable edge points, and eliminating the points with the variance larger than or equal to the set threshold value. Preferably, in step S5, when calculating the distance between two parallel edge lines, a point-to-line distance formula is used for calculation. The invention also discloses a computer readable storage medium and a computer system, wherein the computer readable storage medium is provided with a computer program, and the computer program executes the method according to any one of the above methods after running. A computer system comprising a processor, a storage medium having a computer program stored thereon, the processor reading and running the