Study on fabric density identification based on binary feature matrix

Abstract

In order to correctly identify the fabric density, a new algorithm based on binary feature matrix is proposed in this paper. A best threshold plane is determined by the gray level of fabric image, the 3D gray image of fabric is segmented according to the best threshold plane and a gray image of feature cluster is obtained. Then the feature cluster is normalized to transform the fabric image into a binary feature matrix only with threshold value and feature value. Finally, equations for density computation of fabric with binary feature matrix are given. Experimental results show that the proposed algorithm possesses features of accurate computation and fast speed, which can effectively identify the fabric density and provide a new idea for density recognition.