Signature method for determination of the lunar lander position by video image

Abstract

The paper considers the problem of developing a visual navigation system to determine the proper position of the lunar lander. The paper proposes a new method for comparing the observed image frame with a vector map of the Moon based on the comparison of signatures. Experiments show that the proposed method is able to work in real time, is resistant to lighting conditions, small changes in the camera angle, scale and noise, and is able to work with a large number of gaps and false positives of the crater detector. The proposed method can be used in modern domestic and international space programs for the exploration of the Moon to ensure a soft high-precision safe landing of the lunar lander in a given area of the Moon.