Automatic matching method for remote sensing satellite data ground control point based on knowledge learning

Abstract

The invention provides an automatic matching method for a remote sensing satellite data ground control point based on knowledge learning, comprising the following steps: contour extraction: contour information of the remote sensing satellite data ground control point is obtained according to an image of a template; content extraction: content information of the remote sensing satellite data ground control point is obtained according to image information of remote sensing satellite data; synthesis: the obtained contour information and content information are synthesized to obtain complete information of the remote sensing satellite data ground control point; matching: the obtained complete information of the remote sensing satellite data ground control point is matched with the template toobtain the offset of the remote sensing satellite data ground control point. By adopting the method, errors brought by the attitude of the remote sensing satellite can be accurately estimated, thereby having important application value in the processing work of the remote sensing satellite data.

Description

technical field [0001] The invention relates to a method for automatically matching ground control points of remote sensing satellite data, in particular to a method for automatically matching ground control points of remote sensing satellite data based on a knowledge learning mechanism, and belongs to the field of remote sensing technology. Background technique [0002] At present, remote sensing satellite images have become an important data source for people to obtain ground information. In the process of processing remote sensing satellite data, there are still some problems to be solved. The first of these problems is: where is the observation object remotely sensed by the satellite? That is, what is the corresponding relationship between each pixel in the remote sensing satellite image and the geographic longitude and latitude on the ground. This problem needs to be solved by remote sensing satellite image navigation technology. [0003] The general process of remot...

AI Technical Summary

Problems solved by technology

[0005] However, the above-mentioned prior art still has certain limitations, which are mainly manifested in: 1. When it is necessary to process a large amount of remote sensing satellite data that is not rich in historical data, the automaticity of the ground control point matching method is difficult to satisfy

The requirement for the automaticity of the method is the prerequisite for the application of a large number of remote sensing satellite data; 2. The existing replacement method is to replace it manually or semi-automatically, and then return the obtained landmark offset to the system to obtain the relevant Therefore, the time and accuracy of image navigation calculations will have large deviations due to manual influence; 3. Existing methods require experts to formulate a large number of template libraries. Due to historical data accumulation and other factors, my country's remote sensing Satellite ground detection data is not rich, so the landmark navigation method suitable for foreign remote sensing satellites is not suitable for remote sensing satellites designed by my country

Images