Radar Video Manual Measurement Extraction

In this proposed Phase-I SBIR project, we will investigate, develop, and demonstrate the feasibility of generic techniques and methods for characterizing the errors associated with manual sensor measurement extraction in a variety of processing environments that have manual tracking capability. Our technical approach will be based on a combination of mathematical and statistical techniques to handle manual measurements under several different settings. These techniques will include operator-input calibration testing, bias estimation, target echo extraction, Kalman filtering and smoothing, and a method of information extraction we refer to as pseudo-measurement. We will refer to our proposed collection of algorithmic methods and techniques as the Radar Video Manual Measurement Extraction (RVMME) System. We recognize that this system will need to be configured to the particular processing environment of interest. The Phase-I development and feasibility study will demonstrate the potential effectiveness of the RVMME system and lead to Phase-II development of a working prototype for incorporation into a specific decentralized, distributed sensor track processing environment, to be identified in consultation with the sponsor.