Multi-layered Multi-Sensor Dynamic Detection and Discrimination of Dismounts

With the recent changes in threats and methods of warfighting, Intelligence, Surveillance and Reconnaissance (ISR) activities have become critical to the militarys efforts to maintain situational awareness and neutralize the enemys activities. While a number sensor platforms exist for ISR and many others are currently under development, each sensor alone faces limitations in meeting the ISR critera of world-wide perspective, fusion, detail and persistent surveillance. However, the complementary nature of these sensors facilitates a synergistic approach to Automatic Target Recognition (ATR) and Tracking in ISR scenarios. To address this need, SSCI proposes the Discrimination using Sensor Collaboration and Verification (DiSCVer) system for dismount detection. We propose an architecture that calls for a high altitude, low resolution sensor generating cues that are passed along to available close-in sensing platforms for high resolution coverage of high priority targets. The proposed system will be flexible with regards to available sensors and will build on existing SSCI capabilities to develop efficient algorithms for dismount detection, density estimation, tracking and discrimination. We also propose the use of Learning Classifier Systems (LCS) to fuse features, contextual information and evidential information obtained from different sensors. BENEFIT: The algorithms developed in this effort will be of great use in many Intelligence, Surveillance and Reconnaissance (ISR) systems used by the military in the Global War n Terror and the government for homeland security. The automated algorithms proposed in this effort will help reduce the burden on personnel involved in these applications. In addition, civilian applications such as environmental monitoring, medical engineering, etc.,  where detections algorithms are expected to perform consistently in a number of different operating conditions, can benefit from algorithms developed to fuse information from different sensors.