A New Parameter Method for Identifying Animal Fibers
Scale structure pattern of animal fiber is different and that is a major reference distinguishing them from each other. Four main absolute parameters, including scale diameters, scale interval, scale perimeter and scale area, are usually used to describe the shape feature of cuticle scale of animal fiber. In this paper, cashmere and fine wool fiber sample are check up under light microscope with a magnification of 40×for objective and their color images are captured by CCD camera fixed in light microscope. After Converting the color images to gray-level images, threshold and morphological manipulations are performed on them and the skeletonized binary images only having one pixel wide edges are obtained. Then, these four absolute parameters describing scale patterns are measured from these binary images and a database composed of numerical data of four relative parameters, including scale diameter, scale interval, normalized scale perimeter and normalized scale area, are established. Finally, a multi-parametr Bayes classification model is developed based on the four relative parameters. The numerical experiment results show that this model is more effective than that based on four absolute parameters when distinguishing cashmere from fine wool (70s) and the classification accuracy is higher than 90 percent.
absolute parameter morphological manipulations relative parameter scale pattern Bayes classification model
SHI Xianjun YU Weidong
Textile materials and technology lab, Donghua University, Shanghai 200051, P. R. China Department of Textile materials and technology lab, Donghua University, Shanghai 200051, P. R. China
国际会议
上海
英文
791-794
2007-10-15（万方平台首次上网日期，不代表论文的发表时间）