Affiliation:
1. Beijing Institute of Fashion Technology
Abstract
This paper studied how to identify the cashmere and wool fibers by scale density. The scale density is extracted in frequency domain. The theory of extracting scale density and image preprocessing are given in this paper. The samples are 100 cashmere images and 100 wool images. The recognition rate is above 83% by SVM cross validation. The result reveals that scale density is a valid feature in identifying the cashmere and wool fiber.
Publisher
Trans Tech Publications, Ltd.
Reference5 articles.
1. Yi Haiyan, Research of cashmere feature parameter extraction techniques based on image analysis[D]. Beijing institute of fashion technology, (2009).
2. Jang Gao-ping, Zhong Yue-qi : A recognition method for wool and cashmere fiber based on the character of the spectral line[D]. Dong hua university, (2010).
3. Chih-Chung Chang and Chih-Jen Lin, LIBSVM : A library forsupport vector machines, 2001. Software available at http: /www. csie. ntu. edu. tw/~cjlin/libsvm.
4. Rafael C. Gonzalez, Richard E . Wood, Steven L. Eddins. Digital Image Processing Second Edition.
5. Peng Wei-Liang, Jiang Yao-xing, Yuan Chang –tai. The analysis of shape feature of cashmere and fine wool[J]. shanghai textle science and technology, 2005, pp.33-11.
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献