Image Processing in Optical Guidance for Autonomous Landing of Lunar Probe

TL;DR

This paper presents a computer vision-based image processing algorithm for autonomous lunar probe landing, enabling precise navigation despite communication delays, with efficient feature detection and tracking demonstrated through simulation.

Contribution

It introduces a novel image processing algorithm for lunar probe navigation that improves efficiency and accuracy over existing methods.

Findings

Algorithm reduces computational load

Successfully detects and tracks feature points

Meets navigation accuracy requirements

Abstract

Because of the communication delay between earth and moon, the GNC technology of lunar probe is becoming more important than ever. Current navigation technology is not able to provide precise motion estimation for probe landing control system Computer vision offers a new approach to solve this problem. In this paper, author introduces an image process algorithm of computer vision navigation for autonomous landing of lunar probe. The purpose of the algorithm is to detect and track feature points which are factors of navigation. Firstly, fixation areas are detected as sub-images and matched. Secondly, feature points are extracted from sub-images and tracked. Computer simulation demonstrates the result of algorithm takes less computation and fulfils requests of navigation algorithm.