Advanced Algorithms for Space-Based Next-Generation Infrared Sensor Exploitation

ABSTRACT: Toyon Research Corporation and Argo Science Corporation propose research and development of innovative algorithms for tracking multiple closely-spaced targets in space-based next-generation infrared imagery. The proposed algorithms are capable of automated detection and tracking of both bright and extremely dim targets in wide field of view image sequences based on advanced spatiotemporal clutter rejection, as well as near-optimal track-before-detect processing and multiple target hypothesis management in a Bayesian estimation framework. The proposed algorithms are capable of super-resolution separation of closely-spaced objects, with this processing applied judiciously within the field of view so as to maintain the capability of real-time processing. In Phase I, feasibility demonstration will be performed using physics-based imagery simulation and preliminary algorithm implementations, and analysis of computational requirements and suitable CONOPS will be performed in preparation for development and verification of a real-time prototype implementation in Phase II. BENEFIT: Successful completion of the proposed research and development will result in significant improvements in the ability to efficiently exploit next-generation satellite-based wide field of view EO/IR sensor data. The developed algorithms and software can be directly applied to real-time processing in SBIRS ground stations. The developed technology is also expected to be directly applicable to battle space awareness and technical intelligence, and is also well-suited for transition into the PTSS and SBSS systems. Broader applications include airborne EO/IR surveillance, including in wide area motion imagery (WAMI)/wide area persistent surveillance (WAPS) EO/IR exploitation for systems such as DARPA"s ARGUS-IS and AFRL"s Blue Devil and Gorgon Stare. The proposed research and development will also lay the foundation for development of improved optical detection, tracking, and recognition technology for remote sensing and surveillance applications, as well as for search and rescue, law enforcement, homeland security, and boarder security. The technology could even potentially be introduced in home security systems to provide improved intruder alerting and tracking performance in low-light conditions.