ABSTRACT
Although quality assurance belongs to key factors of profitability in the textile industry, the current inspection processes depend predominantly on human sight. In doing so according to some studies, human visual inspection can catch only around 60 to 75 percent of the defects. Therefore, in order to lower the cost of the inspection process and to increase the competitive advantage of the products, it is necessary to automate this process. In recent years considerable research has been conducted to find robust and fast algorithms on which to base an automated inspection system for textile fabrics. Algorithms under investigation include image analysis, Fourier transform, Gabor filters, Random Markov Fields, and also methods based on wavelet analysis. We provide here short review of these methods.
