Defect pattern recognition on nano/micro integrated circuits wafer

Defects on Nano/Micro Integrated Circuits Wafer (ICW) tend to cluster and the spatial patterns of these defect clusters usually contain important information for quality engineers to remove the root causes of failures. In this paper, a new method consisting of noise filter, defect clustering by using chameleon method and model-based pattern recognition is proposed for automatic defect spatial pattern recognition on Nano/Micro ICW. The new method can not only find the number of defect clusters, and identify the pattern of each cluster, but also provide valuable information for the yield and reliability study. The method can recognize not only the linear/curvilinear patterns, ellipsoidal patterns, but also the ring spatial patterns.