Inspection of IC wafer Defects Based on Image Registration

An automatic integrated circuit (IC) wafer inspection system based on image registration has been developed. This inspection system addresses the task of detection the key defects on the surface of IC wafer. It introduces the method of Speeded Up Robust Features (SURF) image registration. The algorithm is used to achieve the alignment of space position between the standard image and the detected image. Defect feature extraction and analysis techniques are studied regarding the characteristics of detected images of IC wafer. The contour extraction with pattern recognition was applied to classify the defects. This method classify the type of defects based on the number in the Boundary Contour and material increase or decrease. The effectiveness of the method is demonstrated through some samples of IC wafer.