Effective Fabric Defect Classification System using the Fast Fourier Transform by Splitting the Color Channels

This research paper focuses on the detection and classification of fabric defects supported digital image processing techniques. The work is intended to supply the upper speed and accuracy of defect detection than human vision and to seek out the supply of the defects. Aiming at the matter for the detection of fabric defect, this paper uses the strategy of Fast Fourier Transform and basic image pre-processing operations to find the material defects; the algorithmic rule planned during this paper will considerably shorten the time of detection on the idea of guaranteeing the proper detection rate. Open CV library and python programming language are employed for the experiment. The experiment result shows that successful defect detection with a 95% accuracy rate, and it's 50% quicker than human vision in materials density calculations.