Fabric defect detection by digital image processing is inevitably bothered by various noise sources.Denoising algorithms in three ways,i.e.sym4 wavelet default threshold,coif2 wavelet soft threshold and bior.3.7 wavelet soft threshold are used in this paper to denoise fabric defect images,and the denoising effects of original noisy images before and after brightening are compared.The study showes that denoising method based on coif2 wavelet soft threshold is suitable for image denoising of fabric with small defects,and its effect is better after increasing original image brightness.%利用图像处理技术对织物疵点检测的过程中不可避免受到各种噪声源的干扰.采用sym 4小波默认阈值、coif 2小波软阈值和bior.3.7小波软阈值消噪算法对织物疵点图像进行消噪处理,并比较了原始含噪图像增亮处理前后的消噪效果.研究表明:coif 2小波软阈值消噪算法适合对织物小疵点图像进行消噪处理,原始图像经过增加亮度后处理效果更佳.
展开▼