基于二层自适应正交小波的疵点检测

本文提出了基于二层自适应正交小波的织物疵点检测方法。首先介绍了织物纹理图像的二维正交小波变换．在此基础上借鉴Daubechies小波构造过程，阐明了二层自适应正交小波的构造方法，然后对图象进行二层小波分解。分别比较无缺陷图像与待检测图像二层分解后的纬向和径向子图像，得到纬向和径向疵点信息，最后将两个方向上的疵点信息融合．得到检测结果。实验证明该方法是有效的。

Fabric Defect Detection Based on Double-deck Adaptive orthonormal Wavelets

A method of fabric defect detection based on double-deck orthonormal wavelets has been developed in this paper. Firstly this paper introduces 2-D orthonormal wavelet transform and illuminates the way of construction of double-deck adaptive orthonormal wavelets,using the one of Daubechies wavelets for reference. Then the defect image is decomposed double-deckly into horizontal and vertical sub-images, and defect detection is made on result of the decomposition, by comparing the sub-images of sample without defect and the ones of the fabric defect image. The horizontal and vertical results of the defect detection are fused as the final one. The experiment proves the method to be effective.