Enhanced Synthetic Aperture Radar (SAR) Imaging of Moving Targets Using the Deterministic Annealing Probabilistic Multi-Hypothesis Tracker (DAPMHT) Approach

Abstract

Synthetic aperture radar (SAR) imaging of a moving target has been the subject of interest in both military surveillance, reconnaissance, and civilian remote-sensing applications because of its high-resolution target imaging and tracking capability. Current SAR moving target imaging requires that estimates of the target states be obtained before any imaging can be performed. Processing SAR data accurately will produce an efficient image; however, SAR was primarily designed for the imaging of stationary targets. Thus, for imaging moving targets, traditional SAR algorithms suffer from low imaging quality (i.e., blurring), inaccurate target state estimation, and the inability to solve for displacement/defocusing due to the target’s uncertain motion. As a result, degraded target image quality. This dissertation addresses these particular challenges and presents an analysis of the effects of moving targets on SAR imaging. This dissertation aims to resolve the problem of moving target imaging in SAR by combining a tracking algorithm with the SAR algorithm to maximize the overall performance and handle the complexity of SAR imaging. Our proposed processing algorithm combines deterministic annealing probabilistic multi-hypothesis tracking (DAPMHT) and range-Doppler algorithm (RDA) imaging to continuously observe a scene and generate a SAR image simultaneously. This development approach also improves the algorithm’s global convergence. The DAPMHT method is adapted in this dissertation as the basis for a principal and overall study of SAR imaging of moving targets. This work focuses on cases of direct energy returns from point targets and extended targets that are moving along a straight-line path or maneuvering. The proposed algorithm’s advantages are: (i) the final SAR image does not need to be modified because all the target motions are compensated for during the image formation process (ii) both point and extended moving targets can be tracked and imaged without employing different filters for each target and (iii) the algorithm is capable of tracking and imaging moving targets in a dense clutter environment. This dissertation demonstrates and validates the feasibility of the DAPMHT approach. Results and work have been done to accurately estimate the target parameters and focus the moving target images; they show the effectiveness and robustness of combining the tracking DAPMHT algorithm with SAR imaging.