作者: Xingzhi Chang , Chengxi Gu , Jiuzhen Liang , Xin Xu
DOI: 10.1155/2018/3709821
关键词:
摘要: This paper proposes a novel template-based correction (TC) method for the defect detection on images with periodic structures. In this method, fabric image is segmented into lattices according to variation regularity, and applied reduce effect of misalignment among lattices. Also, defect-free are chosen establishing an average template as uniform reference. Furthermore, procedure composed two steps, namely, defective locating shape outlining. Defective based classification patterns, which involves improved E-V centralized processing, while outlining provides pixel-level results by threshold segmentation. we also present some experiments detection. Experimental show that proposed effective.