Fabric defect detection using a genetic algorithm tuned wavelet filter

This paper presents a method to design a wavelet-filter that minimizes entropy in the wavelet transform of images of woven fabrics. Filters that minimize entropy in images tend to filter out fabric texture while highlighting fabric defects. The design of the wavelet filter is couched as a non-convex optimization problem which is solved using a hybridized Genetic Algorithm. Three distinct filters are tuned to detect horizontal, vertical and blob defects in woven fabrics. In addition to texture filtering, defect segmentation, noise removal, and object extraction are presented. The effects of shifting on the optimized set of coefficients is also explored.