Wavelet Entropy Based Floating Thread Defect Detection for Tufted Carpet

Abstract

This paper proposed the floating thread defect detection method base on wavelet entropy for tufted carpet. The method first get the image from cam and preprocesses the captured image with filtering, then using Wavelet Entropy to judge whether there are defects in the image. For image consist of defects, apply further thresholding operation to the image. Closing operation was adopted to eliminate noise and highlight the defects. At last, nine samples image were tested to verify the feasibility of proposed method.