Variable Remapping of Airborne Imagery

UAV to target-platform handoff requires the capability to remap imagery to conditions of the target platform to enable accurate identification (ID). SET proposes to design, demonstrate feasibility, and implement an operational system to provide variable remapping of airborne imagery. We propose to evaluate novel view synthesis solutions for instances where targets lie predominantly on a dominant plane as well as those for which non-planar characteristics are extractable. Rapid methods for both cases using computed homographies, properties of the image acquisition scenario, and platform metadata will be analyzed for effectiveness in terms of view synthesis and real-time implementation. We will analyze performance with respect to system parameters including range and perspective uncertainty, reliability of the synthesized view, latency effects and the overall enhancement of confirmatory ID (CID) due to the view synthesis. The system design incorporates correction for auxiliary conditions, including atmospheric and sensor characteristics, to provide target imagery that accurately depicts the expected view from the targeting platform. Feasibility of the proposed approach will be demonstrated in Phase I, with Phase II focusing on the development of a fully evaluated prototype integrated within the Video Verification of Identity (VIVID) platform.