Classification and identification of cotton foreign fibers based on support vector machine
Cotton foreign fibers have seriously affected the quality of cotton products. The classification and identification of cotton foreign fibers is the foundation of the cotton foreign fibers automated inspection. The paper takes the typical cotton foreign fibers in Chinas textile enterprises as the research object, and acquires the images under the simulated actual cotton processing. The various classification features are calculated and analyzed. The results show that aspect ratio, roundness, duty cycle and I1 are the effective features to classify various foreign fibers. The paper puts forward the classifier of cotton foreign fibers based on support vector machine. Decision Tree Support Vector Machine(DTSVM) can not only avoid the non-separated region, but also improve the train speed while the training sample number gradually decreased along the decision tree. DTSVM is be used to identify the sort of common foreign fibers in cotton. The experimental data shows that the identify rates of different types of foreign fibers are greater than 92％ by using proposed DTSVM.
Cotton Foreign fibers Classification features Support vector machine
Ronghua Ji Daoliang Li Lairong Chen Wenzhu Yang
College of Information and Electronic Engineering,China Agricultural University,Beijing,100083,China School of Technology,Beijing Forestry University,Beijing,100083,China
国际会议
北京
英文
1-6
2009-10-14（万方平台首次上网日期，不代表论文的发表时间）