Automatic woven fabric structure identification by using principal component analysis and fuzzy clustering

The goal of the study is to develop automatic fabric analysis system by using inexpensive image processing techniques. In this study, we proposed a novel automatic method for woven fabric structure identification. This method is based on widely used digital image analysis techniques. It allows automatic weft yarn and warp yarn cross area segmentation through a spatial domain integral projection approach. Secondly, texture features based on grey level occurrence matrix are studied and optimized by applying principal component analysis. The optimized texture features are analyzed by fuzzy c-means clustering for classifying the different cross area states. The texture orientation features are calculated to determine the exact state of cross area. Finally, woven fabric structures, for example, weave patterns and yarn counts are automatically determined. To verify the validity of this method, a number of sample images are used. The samples have different weave types, different fiber appearances and yarn counts. The recognition results match the actual structure of tested samples.