SAR-to-HRR Feature Association for Move-Stop-Move Tracking

"An operational move-stop-move tracking capability would lead to dramatic improvements in overall battlefield awareness because it would increase the availability of sensing and exploitation assets that currently must be tasked to re-establish targetidentity after track breaks. The critical challenge in move-stop-move tracking is to associate a collection of HRR measurements of a moving target with a SAR image of a stationary target. To address this challenge, we propose an innovative stochasticframework for constructing, from a SAR image, a probability density function for observed HRR profiles connected with a target track. This approach makes it possible to extrapolate information in a SAR image collected at a given target aspect to predictHRR profiles at other aspect angles while appropriately accounting for signature uncertainties, which increase as the aspect angle differences increase. The approach also provides a rigorous framework for directly computing the likelihood of an HRRmeasurement, which can be used as a measure of association strength. The probabilistic formulation is important because it supports natural integration into a multi-hypothesis tracker. We propose a systematic, comprehensive plan to characterize therobustness of our association method using real moving and stationary target signatures from the DARPA MTFP program. The technology developed under this program will contribute directly to the overall mil