Welding mark inspection method and apparatus, and electronic device

Patent metadata

Field	Value
Publication number	US-12134144-B2
Application number	US-202318296993-A
Country	US
Kind code	B2
Filing date	Apr 7, 2023
Priority date	Jun 30, 2022
Publication date	Nov 5, 2024
Grant date	Nov 5, 2024

Abstract

A welding mark inspection method includes acquiring a picture including a welding mark, and obtaining a classification type of the welding mark based on the picture and a welding mark classification model. The welding mark classification model is configured to classify the welding mark based on characteristics of the welding mark.

First claim

What is claimed is:

1. A welding mark inspection method, comprising:
acquiring a picture, wherein the picture comprises a welding mark; and
obtaining a classification type of the welding mark based on the picture and a welding mark classification model, wherein the welding mark classification model is configured to classify the welding mark based on characteristics of the welding mark, and obtaining the classification type of the welding mark based on the picture and the welding mark classification model comprises:
in response to the welding mark being an anode welding mark, obtaining the classification type of the anode welding mark based on the picture and a welding mark classification model for the anode welding mark; and
in response to the welding mark being a cathode welding mark, obtaining the classification type of the cathode welding mark based on the picture and a welding mark classification model for the cathode welding mark, the welding mark classification model for the cathode welding mark being different from the welding mark classification model for the anode welding mark.

2. The method according to claim 1 , wherein the characteristics of the welding mark comprise at least one of shape, color, or area of the welding mark.

3. The method according to claim 1 , wherein obtaining the classification type of the welding mark based on the picture and the welding mark classification model further comprises:
obtaining a classification value based on the picture and the welding mark classification model; and
obtaining the classification type of the welding mark based on the classification value.

4. The method according to claim 3 , wherein:
the classification value comprises a first probability, a second probability, and a third probability, wherein the first probability is a probability of the classification type being acceptable, the second probability is a probability of the classification type being burst point, and the third probability is a probability of the classification type being bad; and
obtaining the classification type of the welding mark based on the classification value comprises:
determining a classification type corresponding to a maximum value among the first probability, the second probability, and the third probability as the classification type of the welding mark.

5. The method according to claim 1 , wherein:
the picture is a first picture; and
acquiring the first picture comprises:
acquiring a second picture, wherein the second picture comprises a welding region in which the welding mark is located; and
segmenting the first picture from the second picture based on the second picture and a welding mark segmentation model.

6. The method according to claim 5 , wherein the welding mark segmentation model is configured to segment the first picture from the second picture based on a boundary shape of the welding mark.

7. The method according to claim 5 , wherein acquiring the second picture comprises:
acquiring a third picture, wherein the third picture comprises a surface region in which the welding region is located; and
extracting the welding region in the third picture to obtain the second picture.

8. The method according to claim 1 , wherein the classification type comprises acceptable, burst point, or bad.

9. The method according to claim 8 , further comprising, in a case that the classification type of the welding mark is burst point:
extracting a target region in the picture, wherein the target region is a region with a sudden color change in the picture; and
determining the welding mark as acceptable or bad based on the target region.

10. The method according to claim 9 , wherein determining the welding mark as acceptable or bad based on the target region comprises:
determining the welding mark as acceptable or bad based on the target region and a region classification model.

11. The method according to claim 10 , wherein the region classification model is configured to classify the welding mark based on characteristics of the target region.

12. The method according to claim 10 , wherein determining the welding mark as acceptable or bad based on the target region and the region classification model comprises:
obtaining a classification value based on the target region and the region classification model; and
determining the welding mark as acceptable or bad based on the classification value.

13. The method according to claim 9 , wherein extracting the target region in the picture comprises:
obtaining the target region based on the picture and a region segmentation model, wherein the region segmentation model is configured to segment a region with a sudden color change in the picture as the target region.

14. The method according to claim 9 , wherein determining the welding mark as acceptable or bad based on the target region comprises:
determining the welding mark as acceptable or bad based on at least one of a gray level or an area of the target region.

15. The method according to claim 14 , wherein determining the welding mark as acceptable or bad based on the at least one of the gray level or the area of the target region comprises:
in a case that the area of the target region is greater than or equal to a first threshold, or in a case that the gray level of the target region is greater than or equal to a second threshold, determining the welding mark as bad; or
in a case that the area of the target region is less than the first threshold and the gray level of the target region is less than the second threshold, determining the welding mark as acceptable.

16. A welding mark inspection apparatus, comprising:
a processor; and
a memory storing instructions that, when executed by the processor, cause the processor to:
acquire a picture, wherein the picture comprises a welding mark; and
obtain a classification type of the welding mark based on the picture and a welding mark classification model, wherein the welding mark classification model is configured to classify the welding mark based on characteristics of the welding mark, and obtaining the classification type of the welding mark based on the picture and the welding mark classification model comprises:
in response to the welding mark being an anode welding mark, obtaining the classification type of the anode welding mark based on the picture and a welding mark classification model for the anode welding mark; and
in response to the welding mark being a cathode welding mark, obtaining the classification type of the cathode welding mark based on the picture and a welding mark classification model for the cathode welding mark, the welding mark classification model for the cathode welding mark being different from the welding mark classification model for the anode welding mark.

17. The apparatus according to claim 16 , wherein the characteristics of the welding mark comprise at least one of shape, color, or area of the welding mark.

18. The apparatus according to claim 16 , wherein the instructions further cause the processor to obtain the classification type of the welding mark based on the picture and the welding mark classification model by:
obtaining a classification value based on the picture and the welding mark classification model; and
obtaining the classification type of the welding mark based on the classification value.

19. An electronic device, comprising:
the welding mark inspection apparatus according to claim 17 .

20. A welding mark inspection method, comprising:
ac

Assignees

• Contemporary Amperex Technology Hong Kong Ltd

Inventors

• Chen Fei
• Jiang Guannan
• Chen Xiaoyi

Classifications

• G06T7/0004 Primary

  Industrial image inspection · CPC title

• G06T2207/30152

  Solder · CPC title

• G06T2207/20084

  Artificial neural networks [ANN] · CPC title

• G06T2207/10024

  Color image · CPC title

• G06T7/11

  Region-based segmentation · CPC title