文献类型：会议论文

英文题名：Chinese national forest types identification method using FY-3A MERSI data

作者：Qin, Xianlin[1] Hu, Bo[1] Pang, Yong[1] Li, Zengyuan[1]

第一作者：覃先林

通信作者：Qin, X.|[a0005f9c5255dd62043d8]秦雪;

机构：[1] Institute of Forest Resource Information Technique, Chinese Academy of Forestry, No. 2 Dongxiaofu, Haidian, Beijing, 100091, China

会议论文集：33rd Asian Conference on Remote Sensing 2012, ACRS 2012

会议日期：November 26, 2012 - November 30, 2012

会议地点：Pattaya, Thailand

语种：英文

外文关键词：Decision trees - Forestry - Remote sensing - Timber - Time series

年份：2012

摘要：Forest distribution mapping is one of the direct comprehensive reports of the achievement in forestry survey. To developing the application method in Chinese forestry management and generating Chinese forest cover distribution map at national scale with 250 meter spatial resolution, based on the analyzing results of related bands of FY3A-MERSI by using typical sample, three forest cover mapping methods, including unsupervised classification method (UC), decision tree classification method (DC), and stratified classification method of combining the unsupervised classification result method (SC), have been applied using time series ten days NDVI data of year 2009 of China, which have been generated by using Maximum VI Value Composite (MVC) method. The validation results show that the overall accuracy of UC, DC and SC is 58.30%, 77.46%, and 86.14% respectively. Their Kappa coefficient is 0.5289, 0.7419, and 0.8427 in turn. It shows that the precision and Kappa coefficient of SC is best than other selected two methods.