6‐3: Identifying the Detail Reason of Pixel Defect via Machine Learning Method

Abstract

Various studies have been conducted to automatically inspect defective products. Those technologies already can detect defective pixels better than humans and have been improved enough to detect Mura defect, which is rarely detected by the auto detecting systems. It is as important to determine the cause of the defect as to detect it. However, efforts to automatically identify the cause of the defect have yet to show significant results. In this study, a method to determine the cause of defect with machine learning is introduced.