The application of improved threshold segmentation on detection of color fluff

Abstract

This paper proposes an improved threshold segmentation to turn the image of fluffs into binary image, the noises and impurities in the image are eliminated by using Gauss filtering and the method of features statistics of color fluffs, then the centroid coordinate of each color fluff is calculated and the location information of each color fluff is obtained, finally every color fluff is detected and eliminated. The experiment proved that the threshold segmentation has the characteristics of high detection rate and fast calculate speed.