Detection of fabric weft bar based on cross-correlation function

Due to the effect of disadvantage factors,weft bar are often produced in textile products.The quality of fabric will be debased.A research on detection of weft bar was conducted using cross correlation of random process during image detecting.The relation of the cross correlation value with the similarity of the two images was analyzed.Taking the template image and the sub-image of target image as non-correlation random process,the cross correlation function is described as the inner-relation of two random processes.The bigger the value of the cross correlation function is,the more similar the two images,or vice versa.And verification experiments were implemented.The results indicated that the method is workable and the digitalized grading of the weft bar can be accomplished.