CN-122023254-A - Method and network for detecting dislocation of front and back surfaces of pole pieces not at same position

CN122023254ACN 122023254 ACN122023254 ACN 122023254ACN-122023254-A

Abstract

The invention relates to a method and a system for detecting dislocation of the front and the back of a pole piece at a different position, which are based on an image acquisition device A and an image acquisition device B which are arranged at different positions on two sides of a pole piece in a tape running direction, wherein the image acquisition device A and the image acquisition device B are operated by a same encoder to acquire images of the front and the back of the same pole piece at different tape running positions, then coordinate alignment is realized according to a picture ID difference value between pictures of the same pole piece on different image acquisition devices and coordinate information between characteristic points, an affine matrix is solved through a plurality of characteristic points in a combined way, and finally, the coordinate of the same vertex of the AB side is converted to the same coordinate system according to the affine matrix to realize calculation of dislocation values.

Inventors

ZHANG QUAN
WANG GANG
ZHAO ZHE
FU YANQIAO

Assignees

易鸿智能装备(常州)有限公司

Dates

Publication Date: 20260512
Application Date: 20251226

Claims (10)

1. The utility model provides a positive and negative dislocation detection method of pole piece that is not in same position, based on image acquisition equipment A and image acquisition equipment B of the different positions that set up in pole piece direction of tape-running both sides, both obtain the same pole piece at the A of different tape-running positions, B face image through same encoder manipulation, its characterized in that includes: s10, acquiring an A-plane image and a B-plane image of a correction pole piece, wherein at least three non-collinear characteristic points are arranged on the correction pole piece; S21, extracting feature point coordinates of an A-plane image to obtain an A-plane feature point coordinate set and an A-plane picture sequence number; s22, extracting feature point coordinates of the B-plane image to obtain a B-plane feature point coordinate set and a B-plane picture sequence number; and S30, aligning the coordinates of the feature points of the surface A with the coordinates of the feature points of the surface B according to the coordinates of the feature points of the surface A, the picture sequence number, the coordinates of the feature points of the surface B and the picture sequence number to obtain a first feature point coordinate set of the surface A. S40, solving an affine transformation matrix according to the characteristic points in the first A-plane characteristic point coordinate set and the corresponding characteristic points in the B-plane characteristic point coordinate set; And S50, carrying out coordinate alignment on the vertex coordinates of the pole piece coating region in the A-plane image according to the affine transformation matrix, solving the corresponding coordinates in the B-plane image through the affine transformation matrix, and calculating the distance between the corresponding coordinates and the vertex coordinates of the pole piece coating region of the B-plane image to obtain the dislocation value of the front and back surfaces of the pole piece.
2. The method for detecting dislocation of front and back surfaces of a pole piece not at the same position according to claim 1, wherein the coordinates of any a-plane feature point in step S30 are aligned by the following formula: Wherein, the Is a characteristic point sequence number, A picture number of A face, Is the picture number of the B face, Is the A surface The X-axis coordinates of the individual feature points, Is the A surface The Y-axis coordinates of the individual feature points, Is the B surface The Y-axis coordinates of the individual feature points, Is the pixel width of a complete picture.
3. The method for detecting dislocation of front and back surfaces of a pole piece not at the same position as claim 2, wherein step S40 includes: s401, presetting unknown X-axis rotation parameters, X-axis scaling parameters, Y-axis rotation parameters, Y-axis translation parameters and X-axis translation parameters, and constructing an affine transformation matrix; S402, performing matrix multiplication operation on each characteristic point coordinate in the first A-plane characteristic point coordinate set and the affine transformation matrix to enable the final transformation result to be consistent with the characteristic point coordinate of the B-plane characteristic point coordinate set picB, so as to obtain a multi-element primary equation set; S403, solving a multi-element once equation set to obtain an affine transformation matrix.
4. The method for detecting dislocation of front and back surfaces of a pole piece not at the same position as in claim 3, wherein when the number of the feature points is greater than three, the step S403 is: and solving the overdetermined multi-element primary variance group through a least square method to obtain an affine transformation matrix.
5. The method for detecting dislocation of front and back surfaces of a non-co-located pole piece according to claim 4, wherein step S403 solves parameters of an affine transformation matrix by the following formula: wherein X' is an N X1 vector composed of X-axis coordinates of the B-plane feature point coordinates, Is an N X3 matrix composed of X-axis coordinates of the first a-plane feature point coordinates and correction terms, Is a 3X 1 vector composed of X-axis transform parameters; Wherein, the Is an N x 1 vector composed of Y-axis coordinates of the B-plane feature point coordinates, Is an N x 3 matrix composed of Y-axis coordinates of the first a-plane feature point coordinates and correction terms, Is a3 x 1 vector composed of Y-axis transform parameters.
6. The method for detecting dislocation of front and back surfaces of a pole piece not at the same position as claimed in any one of claims 1 to 5, wherein, The feature coordinate point extraction method in the step S21, S22 is a hough circle finding algorithm.
7. A system for detecting dislocation of front and back surfaces of a pole piece not at the same position, for executing the method for detecting dislocation of front and back surfaces of a pole piece not at the same position as described in any one of claims 1 to 6, comprising: The image acquisition unit is used for acquiring an A-plane image and a B-plane image of the correction pole piece, wherein at least three non-collinear characteristic points are arranged on the correction pole piece; The characteristic point coordinate extraction unit is used for extracting characteristic point coordinates of an input image to obtain a characteristic point coordinate set of a corresponding image and a picture sequence number, wherein the input image is an A-plane image or a B-plane image; and the coordinate alignment unit is used for aligning the coordinates of the feature points of the A plane with the coordinates of the feature points of the B plane according to the coordinates of the feature points of the A plane, the picture sequence number, the coordinates of the feature points of the B plane and the picture sequence number to obtain a first feature point coordinate set of the A plane. The affine transformation matrix solving unit is used for solving an affine transformation matrix according to the characteristic points in the first A-plane characteristic point coordinate set and the corresponding characteristic points in the B-plane characteristic point coordinate set; And the dislocation value calculating unit is used for carrying out coordinate alignment on the vertex coordinates of the pole piece coating area in the A-side image according to the affine transformation matrix, solving the corresponding coordinates in the B-side image through the affine transformation matrix, and calculating the distance between the corresponding coordinates and the vertex coordinates of the pole piece coating area of the B-side image to obtain the dislocation value of the front side and the back side of the pole piece.
8. The non-co-located pole piece misalignment detection system of claim 7 wherein the affine transformation matrix solving unit comprises: The affine transformation matrix initializing subunit is used for presetting unknown X-axis rotation parameters, X-axis scaling parameters, Y-axis rotation parameters, X-axis translation parameters and Y-axis translation parameters and constructing an affine transformation matrix; The equation set construction subunit is used for carrying out matrix multiplication operation on each characteristic point coordinate in the first A-plane characteristic point coordinate set and the affine transformation matrix to ensure that the final transformation result is consistent with the characteristic point coordinate of the B-plane characteristic point coordinate set, thereby obtaining a multi-element primary equation set; And the equation set decoupling subunit is used for solving the multi-element primary equation set to obtain an affine transformation matrix.
9. The non-co-located pole piece misalignment detection system of claim 8 wherein the set decoupling subunit comprises: and solving the overdetermined multi-element primary variance group through a least square method to obtain an affine transformation matrix.
10. A computer device, comprising: at least one memory and at least one processor; the memory is used for storing one or more programs; the one or more programs, when executed by the at least one processor, cause the at least one processor to implement the steps of the non-co-located pole piece front and back misalignment detection method of any one of claims 1 to 6.

Description

Method and network for detecting dislocation of front and back surfaces of pole pieces not at same position Technical Field The invention relates to the field of computer vision, in particular to a method and a network for detecting dislocation of the front side and the back side of a pole piece which are not at the same position. Background The lithium ion battery is a high-capacity long-life environment-friendly battery, and mainly works by means of lithium ions moving between a positive electrode and a negative electrode. Lithium ion batteries have many advantages including high voltage, large specific energy, long cycle life, good safety, small self-discharge, rapid charge, etc. Therefore, the application fields of lithium ion batteries are expanding, and the lithium ion batteries are widely applied to the fields of energy storage, electric automobiles, portable electronic products and the like The production process of the lithium battery comprises stirring, coating, drying, rolling, slitting, laser cleaning, tab welding and winding the battery core, wherein the anode and cathode materials are uniformly coated on metal foil (such as copper foil or aluminum foil) in the coating process to form the core part of the battery, namely the anode and cathode plates. The alignment degree of the coating needs to be calculated on the AB surface of the pole piece, which is a key parameter in the size detection of the winding battery core, and directly influences the safety and energy density of the battery, and a CCD camera is generally used for acquiring an AB surface image of the pole piece for deviation correction detection in the running process after slitting, but the existing deviation detection method can only detect deviation in the direction perpendicular to the running direction of the pole piece, and the deviation in the running direction of the pole piece is difficult to measure because of no reference system, and CCD cameras are required to be simultaneously arranged on two sides of the same position of the running to shoot so as to realize the alignment of a coordinate system in the running direction, but the requirement of compactness of installation space and installation precision becomes high. Disclosure of Invention Aiming at the situation, the invention provides a pole piece front and back surface dislocation detection method which is not in the same position, firstly, based on an image acquisition device A and an image acquisition device B which are arranged at different positions on two sides of a pole piece in a tape moving direction, the two image acquisition devices A and B are operated by the same encoder to acquire images of the same pole piece in different tape moving positions, then coordinate alignment is realized according to the picture ID difference value between pictures of the same pole piece on different image acquisition devices and coordinate information between characteristic points, then an affine matrix is jointly solved through a plurality of characteristic points, and finally, the coordinate transformation of the same vertex of an AB surface is realized under the same coordinate system according to the affine matrix, so that the calculation of dislocation values is realized. The invention is realized by the following technical scheme, on the one hand, the invention provides a method for detecting the dislocation of the front and the back of a pole piece which is not at the same position, which comprises the following steps: s10, acquiring an A-plane image and a B-plane image of a correction pole piece, wherein at least three non-collinear characteristic points are arranged on the correction pole piece; S21, extracting feature point coordinates of an A-plane image to obtain an A-plane feature point coordinate set and an A-plane picture sequence number; s22, extracting feature point coordinates of the B-plane image to obtain a B-plane feature point coordinate set and a B-plane picture sequence number; and S30, aligning the coordinates of the feature points of the surface A with the coordinates of the feature points of the surface B according to the coordinates of the feature points of the surface A, the picture sequence number, the coordinates of the feature points of the surface B and the picture sequence number to obtain a first feature point coordinate set of the surface A. S40, solving an affine transformation matrix according to the characteristic points in the first A-plane characteristic point coordinate set and the corresponding characteristic points in the B-plane characteristic point coordinate set; And S50, carrying out coordinate alignment on the vertex coordinates of the pole piece coating region in the A-plane image according to the affine transformation matrix, solving the corresponding coordinates in the B-plane image through the affine transformation matrix, and calculating the distance between the corresponding coordinates and the vertex coordinates of the