Genetic algorithm based wavelet filter for automatic fabric defect detection

摘要

Visual inspection of fabric defect is one of the oldest & traditional method. Continuously watching the fabric during manufacturing by human being is very boring and tedious job, which results in inaccurate and biased results. Now a day image processing and pattern generation plays very important role in fabric defect detection. Here we are extracting the features of perfect fabric & compare it with defective fabric in order to detect the actual defective portion using wavelet decomposition technique as this technique is very sensitive to line defects. As well as for further enhancement genetic algorithm is implemented for finding out suitable subset of wavelet coefficients. Then finally thresholding, filtering techniques are used for getting final defective portion for analysis.