A unified framework for false alarm reduction using scene context from airborne sensors

This SBIR Phase II project will develop an integrated system that will extract scene and geometric context from aerial videos and utilize it for robust target detection and false alarm mitigation. The key innovation in this effort is to place localized target detection in the context of the overall 3D scene, its constituents, and activities by modeling the interdependence of targets, scene elements, scene and sensor’s geometry, and target-movement patterns. The Phase I of this effort demonstrated the proof of this concept by developing core technologies that included i) extraction of scene elements (such as, roads, vegetation, buildings) from the imagery, ii) parallax modeling, iii) extraction of geometric context (for example, parallax motion field and relative structure), iv) sensor modeling, and v) context based false alarm reduction. The Phase II effort will be focused towards refinement, advancement, and integration of these enabling technologies as well as handling operational issues that include efficient and effective processing of compressed videos, very high resolution imagery, and sensor metadata. In particular, the Phase II effort will include: implementation and refinement of enabling algorithms, development of an integrated system, software and algorithmic optimization, and detailed quantitative and qualitative evaluation of the system and its components.