Prior information-based ellipse fitting optimization method for optical navigation with a short-arc horizon

Abstract

Optical navigation is a key technology in deep-space exploration. Previous studies have struggled to accurately calculate the celestial model in short-arc horizon scenes, greatly reducing navigation accuracy. To address this issue, this article proposes a high-precision short-arc fitting optimization algorithm based on prior shape information. The suggested technique is a two-step estimation process that uses minimal horizon features to solve the optimal model parameters iteratively. First, the random sample consensus (RANSAC) algorithm is applied to compute the prior shape information of celestial bodies and suppress the interference of outliers. Then, an improved ellipse optimization equation is designed to integrate the shape information while ensuring efficiency. Finally, to perform precise short-arc fitting, the model parameters are iteratively optimized using an adaptive trust region. Numerous comparison tests demonstrate that the proposed approach achieves rapid convergence in three steps and requires approximately 15–20% less minimum arc length than Christian’s method.