Abstract: It is an important but time-consuming job to recognize and analyze the fabric weave patterns and texture parameters correctly. A novel approach used image processing technology is proposed for detecting fabric texture directions and recognizing weave patterns. Since wavelet transform is suited for fabric image decomposition and Radon transform is fit for line detection in fabric texture, their excellent performances are put together to figure out line detection in fabric texture and weave pattern. The correlative experimental results show that the method can obtain the accurate measurement, and provide effective foundation for recognizing basic fabric weaves.

Key words: Wavelet transform, Radon transform, Fabric texture detection, Fabric weave analysis

摘要： 织物组织结构及参数的正确识别与分析是一项费时而又重要的工作。该文提出了一种运用图像处理技术检测织物纹理方向和识别织物组织的新方法。把小波变换对织物图像信号良好的分解性能与Radon变换对纹理直线的检测特性结合，运用于织物纹理方向的检测分析及织物组织识别中。实验表明该方法能较准确地测定纹理方向等结构参数，检测结果为正确推断与识别织物所属的基本组织类型提供有效依据。

关键词: 小波变换, Radon变换, 织物纹理检测, 织物组织分析