A fast image matching method for spacecraft autonomous navigation

Abstract

The invention discloses a spacecraft autonomous navigation rapid image matching method. The method comprises steps that (1), SURF characteristic points are extracted, the scale space named as an image pyramid is established, and characteristic descriptors are established; (2), on the basis of the SURF characteristic points acquired in the step (1), a chessboard segmentation model is established; (3), image registering, SURF characteristic points extracted from a source image and a target image are calculated to describe an European distance of a vector, a certain specific threshold is set, if the European distance is smaller than the threshold, matching succeeds; (4), on the basis of random sampling, statistics accuracy of the RANSAC algorithm is presented; and (5), on the basis of error matching point elimination of the RANSAC algorithm, error-matched characteristic points are eliminated according to the result of the step (4). The method is advantaged in that the image registering process can be effectively improved in combination with image segmentation and the random sampling consistency algorithm, computational complexity of the image registering process is low, and accuracy is high.

Description

technical field [0001] The invention relates to the technical field of digital image processing, in particular to a rapid image matching method for autonomous navigation of spacecraft. Background technique [0002] The traditional spacecraft navigation method requires the spacecraft to conduct regular inspections and relies on the support of the ground system. In wartime, the ground system may be destroyed by the enemy and cause the system to be paralyzed. The use of an autonomous navigation system enables the spacecraft to still function normally when ground communications are interrupted. Compared with various other autonomous navigation methods, the autonomous navigation method based on natural landmarks has unique advantages and can be widely used in the near future. [0003] In the natural landmark autonomous landmark navigation system, the landmark is used as the ground reference object. First, collect a large amount of natural landmark information including its geog...

AI Technical Summary

Problems solved by technology

However, there are still the following problems in its application: (1) The computing resources on the star are limited, which leads to the poor real-time performance of image registration. In the case of limited hardware conditions, the calculation of the image registration process must be reduced. (2) Satellite navigation requires high stability, and the problem of low accuracy in image registration will lead to the failure of orbit determination, which will eventually lead to disastrous consequences

Images