Battery weld mark defect detection method, device and medium based on 3D point cloud segmentation

By acquiring point cloud data using a 3D line laser scanning device and performing matrix arrangement and index segmentation, combined with line fitting, the problems of high false detection rate and computational complexity in weld defect detection are solved, achieving rapid and accurate weld defect identification.

CN121904081BActive Publication Date: 2026-06-09SILICON TECH (CHENGDU) CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SILICON TECH (CHENGDU) CO LTD
Filing Date
2026-03-24
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing weld defect detection methods suffer from high false detection rates, complex algorithms, and low point cloud accuracy. In particular, in battery weld defect detection, traditional methods struggle to quickly and accurately identify defects such as broken welds, cracks, and weld beads.

Method used

Point cloud data is collected using a 3D line laser scanning device. By arranging the point cloud matrix and the index correspondence, rapid segmentation and line fitting are performed. Defects are identified by combining distance threshold judgment. The process is simplified to one matrix index transformation, one region segmentation, and one line fitting operation, achieving rapid and accurate detection.

Benefits of technology

It improves the accuracy and efficiency of weld defect detection, enabling rapid identification and differentiation of defect types such as broken welds, cracks, and weld beads, while reducing computational complexity and time consumption.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121904081B_ABST
    Figure CN121904081B_ABST
Patent Text Reader

Abstract

The application discloses a battery welding mark defect detection method and device based on 3D point cloud segmentation and a medium, relates to the technical field of welding mark defect detection, and establishes quick indexing to realize batch data reading and quick regional segmentation by organizing original point cloud data in a matrix form; after any rectangular region is segmented by columns, each column sub-point cloud retains trajectory continuity along a scanning direction, direct line fitting can be performed to construct a local reference line, and calculation amount is reduced; by calculating distance deviation of point clouds in each segmented region to the corresponding reference line, abnormal points deviating from the reference line can be quickly located, the application only needs once matrix indexing conversion, once regional segmentation and once line fitting operation, avoids time-consuming nearest neighbor search, surface reconstruction or voxelization process in traditional point cloud processing, can accurately identify typical defect types such as broken welding, cracks and welding tumors, and effectively improves detection precision and efficiency.
Need to check novelty before this filing date? Find Prior Art