Analysis using Region Detection Techniques for fabric fault Identification

Abstract

The fabric industry is the most revenue generating business in India. Production of the quality fabric is the ultimate goal of the business which requires producing less defective textile for minimizing production cost and time. Generally the assessment and inspection are done manually and it is more time consuming. An automotive and accurate inspection process is needed to reduce error on identifying fabric defects. If the faults are to be identified manually the time consumption is more comparatively less and the percentage of fault identification is deduced to a lower rate. The present work demonstrates that the method can detect the identifying fabric defects like hole, scratch, dirt spot, fly, crack point, colour bleeding. An analysis has been done on this by using different region detection techniques.

Keywords: Edge detection, canny, gaussain detector, harris laplace, Hessian detector.