A Robust Star Identification Algorithm Based on a Masked Distance Map

Abstract

The authors of this paper propose a robust star identification algorithm for a ‘Lost-In-Space’-mode star tracker for lost-cost CubeSat missions. A two-step identification framework and an embedded validation mechanism were designed to accelerate the process. In the first step, a masked distance map is designed to provide a shortlist of stars, and the embedded fast validation process enables the direct output of validated stars before the second step. In the second step, local similarity is utilized to select a set of stars from those shortlisted, and the final validation procedure rejects all unsatisfactory stars. This algorithm can provide reliable and robust recognition even when the captured star images include severe star positioning errors, missing stars and false stars. The proposed algorithm was verified by a simulation study under various conditions. As low-cost star sensors face harsh and unknown environments during deep space CubeSat missions such as asteroid exploration, the proposed algorithm with high robustness will provide an important function.

Hybrid Grid Pattern Star Identification Algorithm Based on Multi-Calibration Star Verification

“…Star sensors [ 1 ] play a crucial role in attitude control systems and are extensively utilized in deep space exploration missions due to their desirable characteristics including no drift, high reliability, and arcsecond-level accuracy in attitude measurement, which surpasses that of other attitude determination devices, such as solar sensors, geomagnetic sensors, and gyroscopes. However, with the emergence of the trend toward satellite miniaturization, the factors of device weight, size, and cost have become increasingly significant.…”

A detection sensitivity method for star sensor based on star point matching

“…The process of detecting stars with star sensors is influenced by many conditions and factors such as stellar magnitude, background radiation, lens, sensors, and noise. Therefore, the key to calculating detection sensitivity is to find an accurate and effective analysis method [8] . At present, the calculation method for the detection sensitivity of star sensors is based on matching the star index output by the star sensor during ground observation in the field with the corresponding navigation star index in the navigation star catalog.…”