A real-time computer vision system for detecting defects in textile fabrics

This paper proposes a real-time computer vision system for detecting defects in textile fabrics. The developments of both the hardware and software platforms are presented. The design of the prototyped defect detection system ensures that the fabric moves smoothly and evenly so that high quality images can be captured. The paper also proposes a new filter selection method to detect fabric defects, which can automatically tune the Gabor functions to match with the texture information. The filter selection method is further developed into a new defect segmentation algorithm. The scheme is tested both on-line and off-line by using a variety of homogeneous textile images with different defects. The results exhibit accurate defect detection with low false alarm, thus confirming the robustness and effectiveness of the proposed system.