基于矩匹配算法的山区影像地形辐射校正方法研究  

A Method Based on Moment Matching Algorithm to Correct Remotely Sensed Image in Rugged Area

作　　者：钟耀武[1] 刘良云[2] 王纪华 阎广建[1] 

机构地区：[1]北京师范大学地理学与遥感科学学院遥感科学国家重点实验室,北京100875 [2]国家农业信息化工程技术研究中心,北京100089

出　　处：《地理与地理信息科学》2006年第1期31-34,39,共5页Geography and Geo-Information Science

基　　金：农业部"农业环境监测"项目(070405);国家科技攻关计划"三种模式管理决策支持系统的构建研究"(2004BA617B-5)

摘　　要：针对山区遥感影像地形辐射校正问题,应用现有模型(如C校正)进行地形辐射校正很难达到理想效果。引入矩匹配算法,利用DEM数据计算坡度、坡向等地形信息,以特定坡度和坡向数据为参考依据,对影像进行地形辐射校正。通过北京房山区SPOT5影像进行试验,表明该方法能在很大程度上消除地形阴影,更好地反映阴影区域的细节信息,同时光谱特性保真程度较好,原模糊的影像区域通过处理基本上达到有效识别地物的要求。Correction of the topographic effect in Remote Sensing is very complex. Topographic effect can＇t be removed successfully with a traditional correction method, such as C correction method. The paper introduced the moment matching algorithm to eliminate the topographic effects. A Digital Elevation Model （DEM） image was selected to obtain the slope and aspect information, and the digital number （DN） values of the image in different slope and aspect ranges were corrected to the pixels in a referenced slope and aspect range. The method presented in this paper was tested by a SPOT5 image in the rugged areas of Beijing. The results showed that the study removed topographic effects successfully and restored the land - surface information in shadow areas efficaciously.

关 键 词：矩匹配 山区影像 地形辐射校正 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]