一种芯片缺陷检测方法及设备
By employing multi-light source stitching image technology and a two-stage defect detection method, the problems of low accuracy and efficiency in LED chip inspection have been solved, achieving automated and efficient chip defect detection, reducing the rate of manual re-inspection, and improving production efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- JUHAOKAN TECH CO LTD
- Filing Date
- 2024-12-26
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for LED chip defect detection suffer from low accuracy and efficiency. In particular, deep learning-based methods struggle to meet the accuracy requirement of 20ppm, and the high rate of manual re-inspection makes fully automated detection impossible.
Multi-light source stitching image technology is used to capture local wafer images from multiple perspectives under different light sources, generate multi-light source stitching images, and perform two-stage defect detection. First, the image of the chip to be inspected is located and cropped, then detailed defect detection is performed, and finally the complete wafer image is restored.
It improves the accuracy and efficiency of chip defect detection, realizes automated inspection of each chip, reduces the rate of manual re-inspection, and improves production efficiency.
Smart Images
Figure CN119936015B_ABST