Satellite-image automatic, geometric and precise correction method

Abstract

The invention discloses a satellite-image automatic, geometric and precise correction method. The method comprises the following steps: (1) a nominal picture of a satellite image is obtained; (2) the nominal picture is processed to obtain the satellite image, and the binarization processing is carried out on the satellite image; (3) the match operation is carried out on a landmark template database after the binarization processing and the satellite image to be matched after the binarization processing, so as to obtain the landmark offset; (4) the result in the step (3) is processed by adopting the quality control measure, so as to obtain a landmark pair that is successfully matched; and (5) the geometric and precise correction is carried out on the nominal picture by adopting the landmark pair that is successfully matched, so as to obtain the revised locating result. With the adoption of the method, the geometric correction result with higher precision can also be obtained without the large amount of careful probing work for the terrain, thereby lowering the running cost of a satellite-image automatic landmark matching system greatly.

Description

technical field [0001] The present invention relates to an automatic geometric fine correction method for satellite images, in particular to an image collected by remote sensing satellites, which first performs automatic landmark matching to obtain landmark offsets, and then uses successfully matched landmark pairs to obtain geometric fine correction The resulting method belongs to the field of remote sensing technology. Background technique [0002] Remote sensing satellite images have become an important data source for people to obtain ground information. At present, for the images collected by remote sensing satellites in the world, the geometric fine correction method based on landmark matching is used to improve the positioning accuracy, so as to improve the application effect of remote sensing satellite quantitative products. [0003] In the early days of the development of remote sensing technology, researchers usually selected landmarks manually for interactive geo...

AI Technical Summary

Problems solved by technology

However, this invention is mainly applicable to satellite remote sensing images of mountains, and its application range is relatively limited

[0006] According to the research results published by Taejung Kim, Tae-Yoon Lee and Hae-Jin Choi in the paper "Landmark Extraction, Matching and Processing for Automated Image Navigation of Geostationary Weather Satellites" (in "Proc. of SPIE" Vol.5657), due to Due to the difference in observation mechanism between geostationary meteorological satellites and polar orbiting meteorological satellites, the automatic geometric fine correction method suitable for polar orbiting meteorological satellites cannot be well applied to geostationary meteorological satellite data

Images