Anomaly Detection and Intelligent Sensor Resource Management
Agency / Branch:
DOD / DARPA
In Phase I, TSC investigated innovative anomaly detection algorithms for application to advanced UAV-based GMTI radars including VADER, AACER and FORESTER. TSC demonstrated a promising technique for reliably detecting and tracking anomalous groups of dismounts in simulated urban / suburban environments that included civilian pedestrians. TSC also developed a technique for estimating target paths when road databases are not available. In Phase II, TSC proposes to enhance and calibrate the target movement and radar models that were implemented in Phase I for simulation testing, and to mature the promising group detection and tracking techniques. TSC will investigate algorithms that use normalcy modeling and pattern recognition to detect a broad range of anomalous behaviors involving vehicles, dismounted troops and vehicle / dismount interactions. TSC will develop techniques for real-time radar resource allocation and EO/IR sensor cueing to provide robust target detection, tracking and ID. TSC will test these algorithms under surveillance conditions that include severe clutter, terrain / building shadowing, and dense pedestrian foot traffic. TSC will evaluate algorithm performance using a mix of simulated and measured radar data and ground truth as available. TSC then plans to conduct a validation experiment by piggybacking on a VADER or other radar flight test.
Small Business Information at Submission:
Corporate Senior Staff
TECHNOLOGY SERVICE CORP.
1900 S. Sepulveda Blvd Suite 300 Los Angeles, CA 90025
Number of Employees: