High Dimensional Feature for Hyperspectral Image Classification

摘要：

　　Making a high dimensional(e.g.,100k-dim)feature for hyperspectral image classification seems not a good idea because it will bring difficulties on consequent training,computation,and storage.In this paper,we study the performance of a high-dimensional feature by texture feature.The texture feature based on multi-local binary pattern descriptor,can achieve significant improvements over both its tradition version and the one we proposed in our previous work.We also make the high-dimensional feature practical,we employ the PCA method for dimension reduction and support vector machine for hyperspectral image classification.The two real hyperspectral image datasets are employed.Our experimental results with real hyperspectral images indicate that the high dimensional feature can enhance the classification accuracy than some low dimensional.

作者: WANG Cailing WANG Hongwei Yinyong Zhang WEN Jia YANG Fan

作者单位: School of computer science,Xian Shiyou University,Xian China gineering University of CAPF,Xian,China Department of electronic and electrical Engineering,University of Strathclyde,UK School of electronics Engineering,Tianjin Polytechnic University,Tianjin,China

会议类型: 国际会议

会议名称: 2018 International Symposium on Water System Operations(ISWRSO 2018)(2018年水资源系统及调度国际研讨会)

会议地点: 北京

会议语种:英文

页码: 1-4

在线出版日期: 2018-10-12（万方平台首次上网日期，不代表论文的发表时间）