New algorithm fabric defects detection

This paper presents a novel scheme for Detection Defects Fabric (DDF) in clothes. This algorithm is relied on Gradient Estimation (GE), Homomorphic Method (HOM) and Cumulative Histogram (CH) for discrimination (GHOMC). The GHOMC depends on enhancement by HOM in addition to pre-processing followed by gradient estimation utilizing Sobel Edge (SE), finally CH for DDF and classification task. After the CH estimation, the difference between CHs with and without defects fabric is estimated for DDF. Simulation results prove that the success of suggested technique in DDF. It is clear that this scheme gives superior for DDF from point view the difference area between the CHs of fabric images with and without defects, entropy, contrast improvement factor, average gradient and the computation time for DDF.