Display Line Defect Detection Method Based on Color Feature Fusion

Display color line defect detection is an important step in the production quality inspection process. In order to improve accuracy of low contrast defects, we propose a display method based on feature fusion. The saliency channels RG|GR and BY|YB were obtained using relative entropy maximum criterion. Then, combined with channel b calculate red-green blue-yellow fusion maps. map maps was by Finally, segmentation threshold calculated according mean standard deviation map. fused binarized segmented obtain binary defects. experimental results show that for multi-background offline algorithm this paper better than 90%, while other mainstreams fail detect. Compared state-of-the-art algorithms, our capable real-time low-contrast detection.