Automatic streak endpoint localization from the cornerness metric
| Autor: | Jonathan Black, Brien Flewelling, Brad Sease |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
020301 aerospace & aeronautics
business.industry Streak Aerospace Engineering 02 engineering and technology 01 natural sciences Image (mathematics) 0203 mechanical engineering Attitude determination Component (UML) 0103 physical sciences Metric (mathematics) Orbit (dynamics) Point (geometry) Computer vision Artificial intelligence business 010303 astronomy & astrophysics Mathematics |
| Zdroj: | Acta Astronautica. 134:345-354 |
| ISSN: | 0094-5765 |
| DOI: | 10.1016/j.actaastro.2017.02.009 |
| Popis: | Streaked point sources are a common occurrence when imaging unresolved space objects from both ground- and space-based platforms. Effective localization of streak endpoints is a key component of traditional techniques in space situational awareness related to orbit estimation and attitude determination. To further that goal, this paper derives a general detection and localization method for streak endpoints based on the cornerness metric. Corners detection involves searching an image for strong bi-directional gradients. These locations typically correspond to robust structural features in an image. In the case of unresolved imagery, regions with a high cornerness score correspond directly to the endpoints of streaks. This paper explores three approaches for global extraction of streak endpoints and applies them to an attitude and rate estimation routine. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect