Robust Registration of Astronomy Catalog

Fan Tian*, Tamás Budavári, Amitabh Basu, Department of Applied Mathematics and Statistics, Johns Hopkins University

Poster

Due to a small number of reference sources, the astrometric calibration of images with a small field of view is often inferior to the internal accuracy. An important experiment with such challenges is the Hubble Space Telescope. A possible solution is to cross-calibrate overlapping fields instead of just relying on standard stars. Following Budavári and Lubow (2012), we use infinitesimal 3D rotations for fine-tuning the calibration but re-formalize the objective to be robust to large number of false candidates in the initial set of associations. Using Bayesian statistics, we accommodate bad data by explicitly modeling the quality which yields a formalism essentially identical to M-estimation in robust statistics. Our preliminary results show great potentials for these methods on simulated catalogs where the ground truth is known.