Combined Appearance-Motion Tracking for PGMTI (UAV-PGMTI)

Period of Performance: 01/27/2006 - 07/31/2011

$2.5MM

Phase 2 SBIR

Recipient Firm

SET Assoc. Corp.
1005 N. Glebe Rd.Suite 400
Arlington, VA 22201
Principal Investigator

Research Topics

Abstract

Automatic detecting, tracking, identification, and cross platform hand-off of moving targets in urban environment from small, jittery aerial platforms are very critical to many military missions in the new century. These tasks are performed poorly today. We will build on Phase I success in development and demonstration of a software system for real time moving target detection and tracking. SET Corporation proposes to develop a new generation of PGMTI trackers that features 1) reliable and accurate sensor-motion compensation and generation of wide area mosaics through matching of point features, line/curve features, and global features; 2) low false negative and high detection rates achieved by using novel temporal variance analysis, factorization analysis based on rigidity evaluation, and motion periodicity analysis for target classification; 3) persistent tracking through appearance modeling and matching; 4) cross platform image registration; 5) cross platform target hand-off based on target motion signatures (periodicity, direction, etc.), temporal coincidence, and target geometry; 6) versatile user selectable informative display modes, including fusing videos acquired from multiple platforms; and 7) a low risk, low cost implementation that does not require any special hardware and can be integrated to current and future UAV sensor processors and Gun Ship workstations.