A system for the automated visual inspection of textiles isdiscussed. The system consists of two main components, (1) theextraction of the texture features utilising the Karhunen-Loeve (KL)transform which provides optimal compression of the image data into afeature vector and (2) the detection of the flaw patterns using aNeyman-Pearson detector, which maximises the rate of detection for aspecified false alarm rate. The performance of the system was evaluatedon various fabrics and different types of textile flaws. The resultsindicate that the system can detect flaws which vary drastically inphysical dimension and nature with a very low false alarm rate.Experimental results in the paper demonstrate the performance of thedetector on some typical textile flaws
展开▼