Robust Feature-Aided Tracking With GMTI/HRR Data

MRLets Technologies, Inc., Raytheon Company and Wright State University will jointly develop a novel GMTI/HRR feature-aided tracking system to vastly improve track continuity for slow moving targets, targets whose velocities below minimum detectablevelocity (MDV) and targets observed intermittently. In the past several years, HRR features have been actively studied as a promising means for extracting information of ground moving targets. However, until now, no research has been reported that canclaim exactly how HRR shall be exploited to achieve the expected performance. Based on our efforts in Phase I, we have identified several bottleneck areas that further study can bring critical breakthroughs in this research. This SBIR Phase II researchaims at maturing the promising developments in Phase I and paving the way for a potential technology transition.We propose to further explore HRR features in a composite format. As there is still no known single HRR feature that can solve the feature-aided tracking problem, several promising different HRR features are considered to jointly assist tracking. Thereare two individual features in the current composite format, namely, the wavelet domain features and the geometric invariants. They will be used in the different phases of a tracking problem. The wavelet domain features will be the main source to supplythe tracking need for continuously observed targets. Model for geometric invariants will be accumulated during this time and used for connecting tracks obtained with discontinued surveillance. As such, advanced tracking requirements such as``finger-printing