Dynamic Vision Sensor for Observing Human-made Space Objects - Detection, Tracking, Characterization
Abstract
More and more human-made objects are launched into the near-Earth and Cislunar space, making detecting, tracking, and characterizing these objects most efficiently an essential task. New sensor technology is required to transcend the current sensing limitations.Neuromorphic, also called neuro-inspired or dynamic vision sensors, are a new type of electro-optical sensor that measures instead of absolute brightness and brightness variations. While neuro-inspired sensors have been very successful in terrestrial surveillance and computer vision, the use for space object tracking and characterization has been severely hampered to the extent of judging the sensor technology unfit for this specific use. This new approach uses classical electro-optical imaging in simultaneous dynamic vision sensors of the DVS type. As a test setup, a dual optic telescope, the Zimmerwald Twin Wide-field Instrument ZimTWIN is used. Ultimately, the groundwork is provided, leading to a sequential design for a single optic. The proposed research is different from previous approaches using either later generations of neuromorphic sensors with triggered imagery or multiple tube designs with different parallel neuromorphic sensors. This new approach of combining classical EO imaging with neuro-inspired imaging allows exact timing for astrometry that has previously been missing for object detection and tracking. For object characterization, the same setup can be used to gain superior temporal resolution exceeding CMOS sensor capabilities without the limitations on the accuracy of the satellite orbits that single photon counters are encountering.At the completion of the project, a new, tested, and validated experimental setup, in addition to a digital twin and methods for neuro-inspired sensors for object detection, tracking, and characterization, is available.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 22, 2024
- Source ID
- FA86552317246
Entities
People
- Thomas Schildknecht
Organizations
- Air Force Office of Scientific Research
- United States Air Force