Garrett Fitzgerald, U.S. Space Force; Zachary Funke, AFRL; Alexander Cabello, Algoritics; Vijayan Asari, University of Dayton; Justin Fletcher, U.S. Space Force
Keywords: WFOV, camera array, GEO detection, GEO tracking, WAMI, wide area motion imagery, SatSim, SatNet, AllSky, space object detection, PANDORA
Abstract:
Persistent, static telescope arrays leverage a scalable imaging architecture to enable detection of dim deep-space objects while maintaining a wide field of view. The PANDORA (Persistent AND Optically Redundant Array) system, a 5×9 array of commercial-off-the-shelf cameras, is an exemplar of this maturing sensing concept located at the Air Force Maui Optical and Supercomputing site. The PANDORA system is designed to capture 20x120deg wide field of view (WFOV) images of the night sky at a rate of two frames per minute, requiring advanced processing strategies to rapidly detect and track objects of interest. Wide area motion imagery (WAMI) object detection methods that extract scene object features are found to be effective in this WFOV space object detection and tracking application, and are used to demonstrate the exploitation capabilities of the PANDORA sensor system. In this work, we apply WAMI-inspired methods to the passive WFOV space object detection problem, with performance measured using a synthetic PANDORA dataset. We report the performance of three WAMI-inspired techniques for geosynchronous equatorial orbit (GEO) object detection, providing baseline GEO object detection measurements in PANDORA imagery.
Date of Conference: September 14-17, 2021
Track: Machine Learning for SSA Applications