A Net Track Solution to Pose-Angular Tracking of Maneuvering Targets in Clutter with HRR Radar

Abstract

Maintaining track continuity is a challenging task for large area persistent surveillance with high target density, rapid target maneuvering, and low target-to-clutter ratio. When the sensor revisit interval is long, measurement-to-track association becomes rather difficult. Although maneuvering targets can be discriminated via Doppler processing, clutter still masks moving targets at certain aspect angles. This creates a blind zone for GMTI (ground moving target indicator) radar, which may extend as large as 25% of the azimuth angles. For high-value targets that may undergo evasive maneuvers, a net track solution coordinating multiple looks from a netted set of sensors is set forth in this paper to ensure continuous monitoring. Target visibility therefore needs to be factored into netted radar platform coordination and sensor systems management. For ground targets, their heading is mostly aligned with the longitudinal axis. The target pose thus carries kinematic information about the direction of velocity vector that can be used to assist tracking of maneuvering targets. In this paper, the range profiles of high range resolution (HRR) radar are used for pose angular estimation of maneuvering targets in clutter. A pose-sensitive classifier provides matching scores that are converted into likelihoods and updated over time in a probabilistic manner. Pose angular tracking from individual sensors in the network is fused to cover the blind zone encountered by individual sensors. A software simulation environment that produces range profiles based on the RF signatures of moving targets is used in the study. Simulation results for pose angular tracking in clutter are presented to illustrate the net track solution concept and its performance.

Estimating target range-Doppler image slope for maneuver indication

“…Among the two possible solutions given by (14), only the one that satisfies the condition of (12) is retained. The slope estimate in pixels is again converted into units of m/Hz by the factor of range pixel spacing over Doppler pixel spacing.…”

“…Furthermore, the HMTI can be used in conjunction with pose angular tracking [14] during HRR profile matching for target ID. This will enhance the target identity confidence on the one hand and to complement the HMTI using differential poses for fast maneuvers missed between scans on the other hand [16].…”

NAECON08 grand challenge entry using the belief filter in audio-video track and ID fusion

“…[23,24] For example, kinematic pose estimates dictate ID feature extraction and ID length-towidth feature estimates determine the target velocity direction. [25] Other examples are Doppler processing in radar [26] and illumination changes in EO [27].…”

Post-update compensation (PUC) with maneuver indicator

“…A direct approach is to estimate the maneuver acceleration [12,13]. One example is to use auxiliary sensors such as an imaging sensor to estimate turns based on rapid change in attitude/orientation following a maneuver.…”