遥感影像的精确配准和正射纠正是进行图像融合、变化检测、图像镶嵌、定量遥感建模、多时相和多传感器影像协同应用的基础和前提。以美国国家航空和航天管理局下设LEDAPS（Land-sat Ecosystem Disturbance Adaptive Processing System）课题组开发的配准与正射纠正程序包AROP（Automated Registration and Orthorectification Package）为例,详细阐述了其配准的原理与程序设计流程,并对其配准的精度进行了分析和评价。试验表明：AROP程序包算法能够找出足够的控制点,且控制点分布较为均匀,配准误差小于0.5个像元。误差特征表现为：扫描误差明显大于航向误差,误差的结果与影像漂移、DEM、坡度存在一定的相关性,高程和坡度是影响配准精度的主要因素之一。该程序包目前能够用于对我国CBERS影像的正射校正以及波段不匹配处理,但是对HJ卫星CCD影像数据配准还有待于进一步研究。
Experiment and Accuracy Analysis of Automated Registration and Orthorectification for Landsat-like Images based on AROP
Accurate registration and orthorectification of remote sensing image is the important fundament for image fusion,change detection,image mosaic,quantitative remote sensing modeling,and synergy application of multi-temporal and multi-sensor images.This paper takes the program package named LEDAPS-AROP（Landsat Ecosystem Disturbance Adaptive Processing System-Automated Registration and Orthorectification Package） of NASA as a case,to illustrate its principle and program design,and analyze its accuracy of registration and orthorectification.The experiment shows that AROP can search enough control points in reference image,and these points distribute evenly.The error of registration and orthorectification is no more than 0.5 pixels,which can meet the requirement of remote sensing quantitative application in mountainous regions.The characteristic of errors presents that the lateral error is obviously bigger than the course error,and the errors seem correlate to the DEM and slope.At present,AROP can be used to process the registration of CBERS image,but still need study further for deal with HJ-CCD image.
NETL
NSTL
万方数据
维普
