Computer Recognition Algorithm for Density Uniformity of Clothing Fabric

摘要

This paper proposes a new computer recognition algorithm which can analyze density uniformity of clothing fabric easily and correctly based on Difference Value Matrix Model (DVMM). Firstly, the clothing fabric image must be pretreated to make it more distinct. Secondly, the clothing fabric image is divided into some segments according to precision demand and a matrix of point's position is built. Thirdly, the weft difference value matrix and the warp difference value matrix are constructed by calculating the difference value between a point and near point. Fourthly, we present Offset Mean Value Formula (OMVF) to analyze density uniformity of clothing fabric for every segment. Fifthly, considering all segments' weft and warp difference value matrix, a changing curve for density uniformity of fabric is depicted. Finally, an example proves that above method is effective for detecting density uniformity of clothing fabric.