遥感多光谱扫描(MSS)数字影象的几何纠正
The Geometric Rectification of Digital Multispectral Scanner(MSS) Imagery in Remote Sensing
-
摘要: 遥感扫描数字影象几何纠正的数学方法,大至可分为随机场的内插法和确定函数场的参数法两大类。就纠正过程而言,又可分为利用可预测的扫描参数作粗纠正和利用控制点作精纠正两大步骤。本文仅介绍用共线法对陆地卫星(Land-sat)多光谱扫描(MSS)影象进行精纠正的基本方法和试验结果。其中,一般多项式内插法被用来确定被纠正象元素的初始坐标;由有限元法产生的数字地面模型(DTM)格网被用于为共线方程提供高程信息。试验结果表明:Landsat MSS影象的几何纠正精度达1~0.5象元单位;共线法比多项式法的纠正精度稍高,但两者无显著差别;由后方交会求解的扫描传感器外方位元素并不代表实际传感器姿态等。本试验还对三种影象灰度的抽样法进行了简单的比较。Abstract: The basic algorithms and experimental results of a precise geometric rectification using collinearity method for Landsat MSS images are introduced in this paper,From the experimental results it is shown,that the rectification accuracy obtained amounts to 1~0.5 pixel unit and no significant accuracy difference exists between the collinear rectification method and the general polynomial one in case of Lands at MSS imagery.The computed exterior elements of the sensor may not stand for its actual attitudes and it is preferable to use bi-linear interpolation method for intensity resampling.