Automation ortho-rectification method based on geometric deformation step resolving

Abstract

The invention provides an ortho-rectification method of an automated medium resolution ratio satellite image, and particularly discloses an automation ortho-rectification method based on geometric deformation step resolving through which an ortho-rectification process of an image can be achieved fast and accurately in a region with obvious topography relief. The ortho-rectification method is built on the basis of analysis of the causes of geometric deformation in the imaging process of a remote-sensing image, different geometric deformation types are separated from one another, and step resolving is conducted to distortion caused by a conforming imaging system and distortion caused by topographic relief. Firstly, simulating irregular deformation caused by topographic relief through imaging simulation, wherein the simulate image serves as an automatic registration reference image, then rectifying confirming imaging system distortion with a polynomial model through an automatic registration method, and lastly resolving topographic deflection existing after target image registration. The automation ortho-rectification method based on geometric deformation step resolving is suitable for images shot by different sensors, other topographic deformation recorded in a lookup table is adopted in topography resolving aiming at different sensors, therefore, the computation speed of the whole rectification process is improved greatly, the ortho-rectification of the images is achieved automatically and effectively.

Description

technical field [0001] The invention relates to a remote sensing image processing technology, in particular to a correction technology for geometric deformation in a remote sensing image and an automatic image registration technology. The invention is suitable for the orthorectification process of the push-broom medium-resolution satellite image. Background technique [0002] Geometric processing of remote sensing images is a key step in preprocessing of remote sensing images, and it is the prerequisite for the correct implementation of subsequent applications such as multi-source image fusion, remote sensing image processing and information extraction. The methods that can be selected for orthorectification mainly include strict physical model and general empirical model. The strict physical model is based on the sensor conformation equation. In order to obtain higher accuracy, the orbital parameters and attitude parameters of the sensor need to be known, and the realizati...

AI Technical Summary

Problems solved by technology

[0005] In terms of satellite image orthorectification process, it can be seen that there are many independent researches in patents / literatures, but there are not many orthorectification solutions proposed from the perspective of the overall process, and they often have differences in automation and local correction accuracy. lacking

Images