地理研究

地理研究 ›› 2008, Vol. 27 ›› Issue (2): 467-477.doi: 10.11821/yj2008020024

• 地球信息科学 • 上一篇    

遥感影像地形校正研究进展及其比较实验

高永年1,2, 张万昌2   

  1. 1. 南京大学国际地球系统科学研究所,南京 210093;
    2. 中国科学院大气物理研究所东亚区域气候-环境重点实验室,北京 100029
  • 收稿日期:2007-03-26 修回日期:2007-10-08 出版日期:2008-03-25 发布日期:2008-03-25
  • 作者简介:高永年(1977-),男,汉族,安徽郎溪人,博士生。主要从事生态水文参数遥感定量化和GIS区域建模研究。E-mail :nianyonggao@163.com *通讯作者 : 张万昌(1966-),博士,"百人计划"研究员,博士生导师。E-mail :zhangwc@tea.ac.cn
  • 基金资助:

    国家重点基础研究发展规划项目 (2006CB400502 )和中科院 "百人计划"择优支持项目(8-057493 )。

Comparison test and research progress of topographic correction on remotely sensed data

GAO Yong-nian1,2, ZHANG Wan-chang2   

  1. 1. International Institute for Earth System Science (ESSI),Nanjing University,Nanjing 210093,China;
    2. Key Laboratory of Regional Climate-Environment Research for Temperate East Asia (RCE-TEA), Institute of Atmospheric Physics, CAS, Beijing 100029,China
  • Received:2007-03-26 Revised:2007-10-08 Online:2008-03-25 Published:2008-03-25
  • Supported by:

    国家重点基础研究发展规划项目 (2006CB400502 )和中科院 "百人计划"择优支持项目(8-057493 )。

PDF (PC)

赞

被引次数

179 | 38

摘要：

地形校正作为复杂地形区遥感影像预处理的重要步骤,对提高地表参数遥感定量化精度具有重要意义。为此,在简述地形校正含义与目标的基础上,回顾并总结了国内外各种地形校正方法并将其划分为基于波段比、DEM和超球面3类方法,以期为地形校正及相关研究提供参考。在DEM的支持下,采用11种地形校正方法对ETM+影像进行了校正比较实验,研究表明:(1)VECA、b、C、Teillet-回归、SCS+C、Minnaert和Minnaert-SCS校正7种地形校正效果较好,可用于遥感影像的地形校正;而Cosine-T、Cosine-C、SCS和Cosine-b校正存在过度校正现象,不宜选择。(2)VECA与b校正模型校正效果最好,且VECA校正比b校正可操作性更强。在此基础上,从地形效应的理论诠释与数学表达、DEM数据、地形校正应用研究3个方面探讨了目前该领域存在的一些问题和难点,并对今后可能的工作重点和研究方向提出了建议。

关键词: 复杂地形, 遥感影像, 地形效应, 地形校正, 研究进展

Abstract:

In remotely sensed image from rugged terrain the sunny surface shows more radiance than the expected and on the shady surface the effect is the opposite. That is to say, a high variation in the reflectance response for the same land use/land cover types is caused by the irregular shape of the terrain. Therefore, the removal of the topographic effect named topographic correction may be critical in areas of rugged terrain, as a preliminary step to the estimation of land surface parameters for Landsat ETM+ image. And then the concept and goal of topographic correction was briefly introduced and then various algorithms for topographic correction developed internationally were reviewed and summarized. Many researches have been carried out on topographic correction and many topographic correction algorithms have been put forward such as cosine correction, C correction, SCS correction, SCS+C correction, Minnaert correction etc. However, there is no clear consensus on methods that may be universally applicable. The main difficulty in applying topographic corrections is related to the lack of standard and generally accepted models. So a correction test was carried out in this study to compare 11 existing popular methods of topographic correction, including a new one named VECA developed by ourselves, and see how successfully they can be applied to Landsat-7 ETM+ image. The test site was selected on the relatively rugged terrain located on the southern piedmont of the Qinling Mountains. And then visual comparison and statistical analysis were adopted for feasible evaluation of the 11 topographic correction methods, and the results suggested that VECA correction, b correction, C correction, Teillet-regression correction, SCS+C correction Minnaert correction and Minnaert-SCS correction have good correction performance, however, the Cosine-T correction, Cosine-C correction,SCS correction and Cosine-b correction would overcorrect the shaded areas in image; and the VECA and b correction are the most capable ones for removing the topographic effects contained in ETM+ image among the 11 methods, compared with b correction, VECA was not only better in performance on topographic effects removal but also simple in theory and easy for operation. In the last section of this paper, the problems and difficulties existing in topographic correction were discussed, and the future hot points and research focuses were discussed.

Key words: rugged terrain, remote sensing image, topographic effect, topographic correction, research progress