所谓遥感数字影像是指通过卫星高空来拍摄地物电磁辐射记录, 以遥感手段获得的影像地图，一般按获取渠道不同分为航片（飞机拍摄）和位片（卫星拍摄）。事实上它是地物的波谱信息、空间信息和时间信息、是地物三维信息的二维投影的综合描述。在遥感影像的成像过程中由于卫星飞行轨道高度、地形起伏、云层厚度等因素,使得拍摄的遥感影像产生一定的几何变形。事实上，遥感影像的几何变形是指像元的坐标系统转换，会在地图坐标系等参考系中，呈现出一定的变化规律。由于这种几何变形会降低遥感图像在应用中的精度，必须用模型修正将几何变形投影到正确的投影系统中，得到高精度的高分辨率卫星遥感影像，提升其正确利用价值。遥感信息处理中最重要的内容之一就是遥感图像的几何纠正。本文提出了多种几何纠正的匹配算法以及纠正精度的分析。 本文对遥感数字影像的性质和特点、处理系统、存储设备、成像模型做了一一介绍。对几何畸变因素、遥感影像几何纠正、地面控制点采集方法和几何精度纠正分析做了深入的研究。通过本文涉及的影像自动匹配算法, 实现了遥感影像几何校正。最终实现遥感影像几何纠正系统的设计与实现。基于四种模块的设计进行开发实践，采用ENVI软件设计技术和软件二次开发来实现几何校正系统。 该论文在实验中采用了人工与自动结合的方法,原始图像和校正图像进行自动匹配，多组控制点采集并选取误差在允许范围之内、分布最为均匀和数量平均的一组控制点数据，进行人工提高控制点精度后，通过两种几何纠正方法，三次重采样，最终获得纠正后的高精度影像。对结果进行分析和总结。本文设计开发了遥感影像几何校正系统，该几何纠正系统采用的是在ENVI软件上进行二次开发,该系统的功能主要有影像数据的存取与输出、显示、控制点的存储、多项式校正和RST几何纠正等，该系统中实现了控制点采集和影像自动几何纠正。最终通过三次重采样得到六组影像数据比对分析，得到相应结论。

The so-called remote sensing digital images via satellite altitude to shoot the feature of electromagnetic radiation records, obtained by means of remote sensing image maps, general access to the channels is divided into different aerial photos (aircraft shooting) and bit-slice (taken by satellites). In fact, it is a feature of the spectral information, spatial information and time information, and feature a comprehensive description of the two-dimensional projection of the three-dimensional information. Of remote sensing images taken from remote sensing image geometric distortion due to the satellite flight track, undulating terrain, thick clouds and other factors, making the imaging process. In fact, the imagery geometric deformation is the image coordinate system conversion, the reference frame of the map coordinate system, showing the difference between the variation. As a result of this geometric distortion will reduce the value of the use of remote sensing images in some applications, you must use the model updating geometric distortion projected onto the correct projection system, high-resolution satellite remote sensing image, so it should correct use of value. One of the most important elements in the remote sensing information processing remote sensing image geometric correction. In this paper, the matching algorithm and correct the accuracy of the analysis of four geometric correction. In this paper, the nature and characteristics of remote sensing digital image processing systems, storage devices, imaging model introduced one by one. Geometric distortion factors, remote sensing image geometric correction, ground control point acquisition method and geometric precision correction analysis to do in-depth research. Automatic matching algorithm, remote sensing image geometric correction by the image of this article. Design and Implementation of the ultimate realization of remote sensing image geometric correction system. Module design based on the four kinds of development practices, using ENVI software design technology and software secondary development to achieve the geometric correction system. Experiment using a combination of methods of doing so with automatic automatic matching, the original image and the corrected image, the plurality of sets of control points to collect and select the error is within the allowable range, the distribution of the most uniform and the average number of a set of control point data, artificially increases the accuracy of the control point, in two geometric correction method, three resampling and, ultimately, high-precision image corrected. The results were analyzed and summarized. In this thesis, design and development of remote sensing image geometric correction system, the system uses the ENVI software secondary development, the functionality of the system image data access and output, display, and storage of the control point, polynomial correction and RST geometric correction etc., the control point acquisition and automatic image matching algorithm is implemented in the system.