Automated Heuristic Defect Classification (AHDC) for Haze Induced Defect Growth Management and Mask Requalification

This article presents results from a heuristic automated defect classification algorithm for reticle inspection that mimics the classification rules. AHDC does not require CAD data, thus it can be rapidly deployed in a high volume production environment without the need for extensive design data management. To ensure classification consistency a software framework tracks every defect in repeated inspections. Through its various image based derived metrics it is shown that such a system manages and tracks repeated defects in applications such as haze induced defect growth.