Space Object Tracking using a Jump Markov System based δ -GLMB filter for Space Situational Awareness

This paper presents a Bayesian filter based solution to the resident space object (RSO) tracking problem using optical telescopic observations. A multiple Probabilistic Admissible Region (PAR) approach, in which multiple hypotheses are made regarding the region of space where Resident Space Objects (RSOs) can appear, is proposed. These hypotheses provide birth target locations within an adaptive birth based δ -Generalized Labeled Multi-Bernoulli (δ -GLMB) filter. This is useful since optical based observations do not provide range information. The dynamics of RSOs is modelled using a Jump Markov System (JMS) within the δ -GLMB filter. Preliminary results show how telescopic images can be processed to generate suitable data for the tracking of RSOs within a JMS δ -GLMB filter.