Defect detection in woven fabric using weighted morphology

This paper aims at investigating a novel solution to the problem of defect detection from the images of woven fabric. Automated visual inspection systems are an attractive alternative to human visual inspection in the textile industry, especially when the quality control of products in the industry is a significant problem. In manual fault detection systems with trained inspectors, very less percentage of the defects are being detected, and thus insufficient and costly. Therefore, automated visual inspection systems are a long felt need in the textile industry. The development of an automated web inspection system requires robust and efficient fabric defect detection techniques. For the detection of fabric defects, the pre-processed image is decomposed into its bit planes. The lower order bit planes are found to carry significant information of the location and shape of defects. Then we find the exact location by means of weighted morphology. The algorithm has been evaluated on a subset of TILDA¹ image database with various visual qualities. Robustness with respect to the changes in parameters of the algorithm has been examined. The test results obtained exhibit accurate defect detection with low false alarms, thus showing the effectiveness and robustness of the proposed detection scheme.