Research and Application of Defect Detection Method for Electronic Wafer Based on Machine Vision

Wafer surface defects have a strong correlation with the manufacturing process in the production process. The accurate and efficient detection of electronic wafer diagrams will certainly improve production efficiency. Based on machine vision and combining traditional image processing with modern information technology, a scheme and framework for defect detection of electronic wafer diagram are proposed. For the Scratch with relatively low detection rate, the parameter feature measurement of Hough transform is proposed, which reduces the false detection rate to some extent. The effectiveness of the proposed model and method is proved by experiments, which has reference value for further application in the production line of IC semiconductor industry.