A Highly Efficient Attitude Estimation Algorithm for Star Trackers Based on Optimal Image Matching

This paper presents a novel attitude estimation algorithm for spacecraft using a star tracker. The algorithm is based on an efficient approach to match the stars of two images optimally on top of each other, hence the name of the algorithm: AIM (Attitude estimation using Image Matching). AIM proved in tests to be as accurate and robust as the existing robust methods, such as q-Davenport, and faster than the fast iterative methods such as QUEST. While this is an improvement in itself, the greatest merit of AIM lies in the fact that it simplifies and in most cases allows to eliminate a very computationally intensive coordinate conversion which normally precedes the attitude estimation algorithm. The computational cost of this conversion step is several times higher than that of the attitude estimation algorithm itself, so this elimination yields a huge increase in efficiency as compared to the existing algorithms. This significant reduction in computational cost could allow to obtain the attitude estimates at a higher rate, implement more accurate centroiding algorithms or use more stars in the attitude estimation algorithms, all of which improve the performance of the attitude estimation. It could also allow the use of star trackers in the expanding field of small satellite projects, where satellite platforms have limited computational capability.