Unsupervised segmentation of printed fabric patterns based on mean shift algorithm

Computer-based fabric segmentation has recently increased significantly in diverse fields of which textiles design and engineering is no exception. Extracting appropriate information from printed fabric manually for further applications has yielded little result due to limitations of the human eye to grasp intricate and complex features. Presented in this study is an unsupervised approach of segmenting printed fabrics using the classical mean shift algorithm by inputting enhanced ranges of parameters based on different number of clusters in the printed fabrics. First and foremost, in order to make printed fabrics more adaptable to the experiment, samples were washed making sure dirt or stains were thoroughly removed, then ironed, scanned and pre-processed eventually with median filter. The filtered samples were then segmented by Mean Shift algorithm. Systematic analysis of experimental results demonstrated that the proposed method is suitable for segmenting printed fabrics with less complexity in clustering.

Automated summary

This study proposes an unsupervised approach of segmenting printed fabrics using the classical mean shift algorithm by inputting enhanced ranges of parameters based on different number of clusters in the printed fabrics. The experimental results demonstrate that the proposed method is suitable for segmenting printed fabrics with less complexity in clustering.