Answers are grounded in this paper and cite their sources.
Accurate star identification remains challenging for star sensors operating under high-maneuvering conditions, where rapid attitude motion degrades stellar geometry and significantly increases the likelihood of misidentification. These difficulties are further aggravated in rolling-shutter exposure modes, in which line-by-line readout introduces additional geometric distortion that compounds the effects of high angular rates. Consequently, the probability of identification failure rises substantially. To address these challenges, a novel, to our knowledge, high-dynamic star identification algorithm for rolling-shutter star sensors is proposed, founded on an optical flow consistency constraint. The key insight is that all correctly matched stellar triangles must yield nearly identical angular-velocity estimates. Leveraging this physical property, the algorithm performs RANSAC-based consensus maximization, followed by an explicit enforcement of the zero z -axis rotation constraint ( ω z =0) and a median-filter refinement. This pipeline enables robust angular-velocity recovery from severely distorted single-frame images. Finally, the experiments demonstrate that the proposed method achieves an identification success rate of 95.2% at 15°/s, demonstrating its effectiveness under rolling-shutter operation. At low angular rates (≤5 ∘ /s), the success rate reaches 99.8%, thereby substantially extending the practical dynamic range of rolling-shutter-based star sensors.
No takes yet. Share an insight, caveat, or question.
Shi et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: