西安工程大学学报2017，Vol.31Issue(3)：383-389,7.DOI:10.13338/j.issn.1674-649x.2017.03.016
基于改进迭代匹配滤波的织物疵点检测
Fabric detect detection based on improved iterative match filter algorithm
摘要
Abstract
In order to detect fabric defects effectively and accurately,a new method based on improved iterative match filter algorithm is presented.Firstly,the fabric image samples are comparatively preprocessed by multiple filtering methods to reduce the Gaussian noise in the sample image and the influence of background texture on fabric defects.Meanwhile,the mean sampling algorithm is determined as the optimal preprocessed result according to objective peak signal to noise ratio and subjective visual effect of the denoised image.Then the processed images are filtered and segmented by improved iterative matching filter.Finally,the contour of the segmented images are smoothed,the defect and isolated points are removed using morphological operation.Experimental results show that the presented method can not only effectively preserve the edge details of the image,but also detect a variety of the fabrics defects.In addition,it has high accuracy and runs fast.关键词
疵点检测/织物瑕疵/迭代匹配滤波Key words
fabric defect detection/fabric defect/improved iterative match filter分类
信息技术与安全科学引用本文复制引用
杨曼,李仁忠,刘阳阳,景军锋,李鹏飞..基于改进迭代匹配滤波的织物疵点检测[J].西安工程大学学报,2017,31(3):383-389,7.基金项目
中国纺织工业联合会科技指导性项目(2013066) （2013066）
西安工程大学大学生创新创业训练计划项目(2016097) （2016097）
