Optimal ring Gabor filter design for texture defect detection using a simulated annealing algorithm

Inspection of small surface defects in texture materials such as fabric and leather is a crucial but challenging task due to high variation of texture features. In this paper, a novel approach is presented that addresses the problem by using an optimized ring Gabor filter (RGF). The parameters of the filter is determined by a simulated annealing (SA) algorithm, which allows the filter to be tuned to a specific range of frequency and orientation so that it can match with the textural features of a defect-free template image. The performance of the proposed method has been extensively evaluated by a variety of real samples with different defect type, shape, size and texture background. The experimental results demonstrated that the scheme is effective and robust, and hence, suitable for real-time inspection applications.