森林资源监测Spot5遥感图像几何精校正方法研究  

Calibration method for forest resource monitoring of Spot5 remote sensing image

作　　者：李崇贵[1] 李春干 

机构地区：[1]深圳职业技术学院3S研究中心 [2]广西林业勘测设计院

出　　处：《北京林业大学学报》2004年第5期22-25,共4页Journal of Beijing Forestry University

基　　金：国家自然科学基金资助项目 ( 3 0 3 7115 9) .

摘　　要：为探索空间分辨率为 2 5m的Spot5遥感图像用于森林资源监测的可行性 ,需进行遥感图像几何精校正 .该文在理论分析遥感图像校正模型的基础上 ,借助GPS虚拟网络参考站定位系统 ,以cm级精度测定校正遥感图像的地面控制点 ,通过实例计算 ,详细研究了在通用遥感图像处理软件ENVI、ERDAS及PCI中各种可能的几何精校正方法及实际精度状况 .综合考虑几何位置变换精度、灰度重采样精度及不同校正方法对地面控制点数量和分布的要求 ,结合林区Spot5遥感图像的特点和GPS定位现状 ,确定了适合森林资源调查的Spot5遥感图像几何精校正方法 .In order study the probability of Spot5 remote sensing image with a spatial resolution of 2 5 m when used to monitor forest resource, calibration of remote sensing image needs to be carried out. On the basis of analyzing calibration model of remote sensing image from theory, by means of GPS dummy network reference station positioning system, the ground control points used for calibrating remote sensing image are surveyed with the precision of centimeter level. According to real example calculation, all possible calibration methods and their real precision are studied in universal remote sensing image processing softwares such as ENVI,ERDAS and PCI. By synthetically thinking of the precision of geometry transform, the precision of gray resample and the requirement to the number and distribution of ground control points, and by integrating the characteristics of Spot5 remote sensing image in forest area and the status of GPS positioning, the calibration method of Spot5 remote sensing image appropriate for forest resource survey is determined.

关 键 词：图像变形 相似变换 多项式变换 灰度重采样 

分 类 号：S771.8[农业科学—森林工程]