On-demand inspection recipe to detect defects of interest using Mahalanobis distance

An on-demand inspection recipe-setup method to detect defects of interest (DOI) was proposed. The method applies Maharanobis distance to recognize DOI-like defects without its own recipe. Moreover, actual application was evaluated and the method effectiveness was confirmed from viewpoints of on-demand processing time and DOI detection. The proposed method enables inspection tool managers to rapidly select an appropriate recipe which detects the most DOI from several initial recipes. The future research work will focus on several examinations for threshold decision based on reviewing all detected defects experimentally.