Supporting Space Domain Awareness Through the Development and Analysis of Space Object Detection Algorithms Employed by Ground-Based Telescopes
Abstract
Detection algorithms are instrumental in maintaining space domain awareness, specifically in the observation, monitoring, and categorization of unknown space objects. State of the art detection algorithms utilize a matched filter or a spatial correlator on long exposure image data to make pixel-wise detection decisions. This thesis investigates the advantages and practical potential of two different post-processing detection algorithms that can be employed by ground-based telescopes. The first algorithm explored is based on a long exposure Fourier domain processing technique, while the second is centered around frame selection from a series of short exposure images. The results of the experiments performed in this thesis ultimately showed that the Fourier point detector algorithm did outperform a traditional point detector algorithm but had significantly lower probability of detection across all false alarm rates when compared to a spatial correlator algorithm over a series of test scenarios. The novel frame selection algorithm was found in the simulated experiment to outperform both the old frame selection algorithm and the spatial correlator in all testing environments at low false alarm rates. The experimental data results confirmed the increased performance of the new frame selection algorithm against its counterparts.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2021
- Accession Number
- AD1134244
Entities
People
- Connor A Paw
Organizations
- Air Force Institute of Technology