摘要
为了克服传统人工目测劳动强度大、误检率高、漏检率高以及现有疵点检测方法检测精度低的问题,提出一种基于正交小波分解的织物疵点检测方法,以提取的特征值相对误差大小代替归一化相对特征值作为判断疵点是否存在的标准。实验结果表明,此种方法比传统方法检测精度高。
In order to overcome such problems as heavy labor intensity,high false detecting rate, high misdetection rate of and low detection precision,caused by the traditional manual visual test,a method of the fabric defect detection,based on orthonormal wavelet decomposition,using the relative error of the extracted characteristic value as the standard for judging the defect instead of the relative character-istic value normalization,is described in this paper.The experimental results proves that the method has a higher detection precision than the traditional one.
出处
《微处理机》
2015年第3期46-49,53,共5页
Microprocessors
关键词
正交小波分解
疵点检测
相对误差
Orthonormal Wavelet Decomposition
Fabric Defect Detection
Relative Error
