Synthesis of Disparate Optical Imaging Data for Space Domain Awareness

Abstract

We present a Bayesian algorithm to combine optical imaging of unresolved objects from distinct epochs and observation platforms for orbit determination and tracking. By propagating the non-Gaussian uncertainties we are able to optimally combine imaging of arbitrary signal-to-noise ratios, allowing the integration of data from low-cost sensors. Our Bayesian approach to image characterization also allows large compression of imaging data without loss of statistical information. With a computationally efficient algorithm to combine multiple observation epochs and multiple telescopes, we show statistically optimal orbit inferences.