Granted Patent B2

US 12,614,399 · App. 18/387,475 · Granted Apr 28, 2026

Inventors: Yinhang Tu (Ningde, CN); Kunpeng Cui (Ningde, CN); Yifei Gao (Ningde, CN)

Assignee: CONTEMPORARY AMPEREX TECHNOLOGY (HONG KONG) LIMITED

G06V20/698G06V10/44G06V20/695

Abstract

Claims (87)

1 . A method for inspecting a cell crush, comprising:

acquiring images of a flowing cell assembly by an image acquisition device;

extracting image features from the images of the cell assembly;

performing similarity matching between the image features and a preset crush model to obtain a similarity matching result, wherein

the preset crush model corresponds to crush feature parameters comprising at least one of gray-scale value, shape, angle, length and width, and scaling, and

the performing similarity matching includes:

identifying, from the image features, similar crush feature parameters that are within different range values of crush feature parameters corresponding to the preset crush model;

determining similar weights corresponding to the similar crush feature parameters based on the similar crush feature parameters; and

obtaining the similarity matching result based on the similar crush feature parameters and the determined similar weights; and

obtaining a cell crush inspection result based on the similarity matching result.

2 . The method according to claim 1 , wherein the acquiring images of the flowing cell assembly by the image acquisition device comprises:

acquiring, by a line scan camera, the images of the cell assembly which flows from a cell assembly station to a pack Mylar station.

3 . The method according to claim 1 , wherein the performing similarity matching between the image features and the preset crush model to obtain the similarity matching result comprises:

obtaining a database of crush feature parameters corresponding to the preset crush model;

obtaining inspection feature parameters of the cell assembly based on the image features; and

performing, in the database of crush feature parameters, similarity matching on the inspection feature parameters to obtain the similarity matching result.

4 . The method according to claim 3 , wherein the performing, in the database of crush feature parameters, similarity matching on the inspection feature parameters to obtain the similarity matching result comprises:

extracting the different range values of crush feature parameters from the database of crush feature parameters;

identifying, from the inspection feature parameters, similar crush feature parameters that are within the different range values of crush feature parameters;

extracting, from the database of crush feature parameters, similar weights corresponding to the similar crush feature parameters; and

obtaining the similarity matching result based on the similar crush feature parameters and the similar weights.

5 . The method according to claim 1 , wherein before the extracting image features from the images of the cell assembly, the method further comprises:

pre-processing the images of the cell assembly, the pre-processing comprising at least one of denoising, image enhancement, gray-scale processing, binary image processing, edge detection and segmentation, and histogram matching.

6 . The method according to claim 1 , wherein before the extracting image features from the images of the cell assembly, the method further comprises:

obtaining sample images labeled with different types of crush features; and

training an AI feature extraction algorithm with the sample images to obtain a trained AI feature extraction algorithm; and

wherein the extracting image features from the images of the cell assembly comprises:

extracting the image features from the images of the cell assembly using the trained AI feature extraction algorithm.

7 . A computer device, comprising:

a memory; and

one or more processors, wherein the memory stores computer-readable instructions that, when executed by the one or more processors, cause the one or more processors to perform the steps of the method according to claim 1 .

8 . One or more computer storage media storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of:

the method according to claim 1 .

9 . The method according to claim 1 , wherein:

the image acquisition device acquires the images of the flowing cell assembly without stopping the cell assembly on a production line.

10 . An apparatus for inspecting a cell crush, comprising:

image acquisition circuitry configured to acquire images of a flowing cell assembly by an image acquisition device;

feature extraction circuitry configured to extract image features from the images of the cell assembly;

similarity matching circuitry configured to perform similarity matching between the image features and a preset crush model to obtain a similarity matching result, wherein

the preset crush model corresponds to crush feature parameters comprising at least one of gray-scale value, shape, angle, length and width, and scaling, and

the similarity matching circuitry is further configured to:

identify, from the image features, similar crush feature parameters that are within different range values of crush feature parameters corresponding to the preset crush model;

determine similar weights corresponding to the similar crush feature parameters based on the similar crush feature parameters; and

obtain the similarity matching result based on the similar crush feature parameters and the determined similar weights; and

inspection circuitry configured to obtain a cell crush inspection result based on the similarity matching result.

11 . The apparatus according to claim 10 , wherein:

the image acquisition device acquires the images of the flowing cell assembly without stopping the cell assembly on a production line.

12 . The apparatus according to claim 10 , wherein the similarity matching circuitry is further configured to:

obtain a database of crush feature parameters corresponding to the preset crush model;

obtain inspection feature parameters of the cell assembly based on the image features; and

perform, in the database of crush feature parameters, similarity matching on the inspection feature parameters to obtain the similarity matching result.

13 . The apparatus according to claim 12 , wherein the similarity matching circuitry, when performing, in the database of crush feature parameters, similarity matching on the inspection feature parameters, is further configured to:

extract the different range values of crush feature parameters from the database of crush feature parameters;

wherein the identifying of the similar crush feature parameters comprises identifying, from the inspection feature parameters, similar crush feature parameters that are within the extracted different range values of crush feature parameters; and

wherein the determining of the similar weights comprises extracting, from the database of crush feature parameters, the similar weights corresponding to the similar crush feature parameters.

14 . A system for inspecting a cell crush, comprising a line scan image acquisition device and inspection circuitry,

wherein the line scan image acquisition device is configured to:

acquire images of a flowing cell assembly, and send the images of the cell assembly to the inspection circuitry; and the inspection circuitry is configured to perform crush inspection by using a method for inspecting the cell crush, the method comprising:

acquiring the images of the flowing cell assembly by the line scan image acquisition device;

extracting image features from the images of the cell assembly;

performing similarity matching between the image features and a preset crush model to obtain a similarity matching result, wherein

the preset crush model corresponds to crush feature parameters comprising at least one of gray-scale value, shape, angle, length and width, and scaling, and

the performing similarity matching includes:

identifying, from the image features, similar crush feature parameters that are within different range values of crush feature parameters corresponding to the preset crush model;

determining similar weights corresponding to the similar crush feature parameters based on the similar crush feature parameters; and

obtaining the similarity matching result based on the similar crush feature parameters and the determined similar weights; and

obtaining a cell crush inspection result based on the similarity matching result.

15 . The system according to claim 14 , wherein:

the line scan image acquisition device acquires the images of the flowing cell assembly without stopping the cell assembly on a production line.

16 . The system according to claim 14 , wherein:

the line scan image acquisition device is configured to acquire the images of the cell assembly as the cell assembly flows from a cell assembly station to a pack Mylar station.

17 . The system according to claim 14 , wherein the inspection circuitry, when performing the similarity matching, is further configured to:

obtain a database of crush feature parameters corresponding to the preset crush model;

obtain inspection feature parameters of the cell assembly based on the image features; and

perform, in the database of crush feature parameters, similarity matching on the inspection feature parameters to obtain the similarity matching result.

18 . The system according to claim 17 , wherein the inspection circuitry, when performing, in the database of crush feature parameters, similarity matching on the inspection features parameters, is further configured to:

extract the different range values of crush feature parameters from the database of crush feature parameters;

wherein the identifying the similar crush feature parameters comprises identifying, from the inspection feature parameters, similar crush feature parameters that are within the extracted different range values of crush feature parameters; and

wherein the determining the similar weights comprises extracting, from the database of crush feature parameters, the similar weights corresponding to the similar crush feature parameters.

19 . The system according to claim 14 , wherein the method further comprises, before the extracting image features from the images of the cell assembly:

pre-processing the images of the cell assembly, the pre-processing comprising at least one of denoising, image enhancement, gray-scale processing, binary image processing, edge detection and segmentation, and histogram matching.

20 . The system according to claim 14 , wherein:

the method further comprises, before the extracting image features from the images of the cell assembly:

obtaining sample images labeled with different types of crush features; and

training an AI feature extraction algorithm with the sample images to obtain a trained AI feature extraction algorithm; and

wherein the extracting image features from the images of the cell assembly comprises:

extracting the image features from the images of the cell assembly using the trained AI feature extraction algorithm.

Assignments (2)

Priority Claims (1)

Continuity (2)

References Cited (25)