Tracking moving weak objects in celestial image sequences

Zhen Lei, Mi Wang, Ting L. Lei, Deren Li

Abstract: One challenge to improving the performance of celestial navigation is the weakness of signals from orbit-known asteroid-like objects. We propose a method for tracking moving asteroid-like objects with weak signals that uses image sequences rather than single images. We employ weak-object enhancement, background subtraction, and multitarget tracking to exploit intraimage, interimage, and object-motion information. Our results demonstrate full tracking without false alarms on a real Laser Geodynamics Satellite image sequence, demonstrating the stability and potential application of the proposed method.