A Robust Contrast Enhancement Method for Automatic TFT-LCD Module Inspection

摘要

An automated detection of defect in TFT-LCD manufacturing is critical to maintain high-quality TFT-LCD production. However, various types of defect and their low SNR make it challenging, and therefore fully automated defect detection is still unresolved. In this paper, we propose a robust method for defect detection combining high-resolution image acquisition technique and image processing techniques. The proposed method shows significantly improved performance in defect detection enough to detect under-gray (≤1gray) defect.