对地观测中心研究生高丛珊获得第五届WSANE最佳论文奖
-
张红 (中国遥感卫星地面站)
-
创建于 2009-11-20
-
2155
2009年11月2日至4日,第五届国际空间航空导航电子学国际会议(WSANE)在上海复旦大学召开。该会议旨在为空间航空导航电子学领域内的一流科学家和工程师,架设起一个关于技术与应用最新进展的国际学术交流平台。本次会议由IEICE协会主办,复旦大学承办,IEEE北京分会、日本JAXA和NICT等共同协办。会议共设立最佳论文奖、最佳学生论文奖、最佳报告奖各1名,由国际电子信息和通信工程师协会颁发。其中,最佳论文奖由对地观测中心在学研究生高丛珊获得。该同学导师为对地观测中心专家张红,王超。
关于WSANE2009的相关信息:
http://www.ieice.org/cs/sane/WSANE2009/award.html
http://www.fudan.edu.cn/fudannews/2009/1109/22746.html
获奖论文的摘要:
SAR change detection based on Generalized Gamma distribution divergence and auto-threshold segmentation
GAO Cong-shan1 2, ZHANG Hong1*, WANG Chao1
1.Center for Earth Observation and Digital Earth, CAS, Beijing, CHINA, 100086
2.Graduate University of Chinese Academy of Sciences, Beijing, CHINA, 100049
(serein_mars@163.com, hzhang@ceode.ac.cn, cwang@ceode.ac.cn)
Abstract
This paper first utilizes generalized Gamma model to fit the statistical characteristics of co-registered SAR images. Then, extract the difference map through measuring the similarity between PDFs by Kullback–Leibler Divergence. Afterwards, apply a combination of KS & KL test to evaluate and choose fitting function of the difference map for the model-based KI threshold segmentation automatically. Experiment was carried on the multi temporal SAR images for Beijing, acquired by TerraSAR and Radarsat-2, as well as Shunyi District of Beijing, acquired by Envisat-ASAR. Such results confirmed the proposed method is more sensitive in detecting regions that have no variance in mean value, but differ in texture, than traditional methods
Keywords: Kullback–Leibler Divergence, Generalized Gamma, KI, Change Detection
责任编辑:张红