文献类型：期刊文献

中文题名：基于小波分析的皮革疵点机器视觉检测方法

英文题名：The Method of Leather Defects Visual Detection Based On Wavelet Analysis

作者：余亚东[1];于梅[1];叶森钢[1]

机构：[1]绍兴文理学院数理信息学院

年份：2010

期号：19

起止页码：232

中文期刊名：微计算机信息

外文期刊名：Microcomputer Information

语种：中文

中文关键词：小波变换;疵点检测;遗传算法

外文关键词：wavelet transform; defect detection; genetic algorithm

中文摘要：提出了一种利用小波变换检测皮革疵点的方法。为了得到能够将疵点区别于正常皮革纹理的小波基,本文使用遗传算法对小波基优化,通过选择适当的适应度函数,得到最佳小波滤波器,使得疵点皮革图像经小波变换后突出疵点信息。实验表明,该方法能够正确检测出皮革图像中的疵点。

外文摘要：A method of leather defect detection based on wavelets transform has been introduced in this paper. For obtaining the wavelet transform coefficients distinguishing defect from natural texture,wavelet transform coefficients were optimized by genetic algorithm. Selecting appropriate fitness function,optimized wavelet filter was designed,then defects were detected by using wavelet transform. This method was proved to be efficient by the experiments.