Improved Frame-Selection Scheme for ISAR Imaging of Targets in Complex 3-D Motion

When a target is in complex three-dimensional (3-D) motions, inverse synthetic aperture radar (ISAR) images cannot be obtained with any motion compensation algorithms. To address this problem, phase nonlinearity can be used to select imaging frames where target 2-D motion is dominant enough to obtain focused ISAR images. Conventionally, adaptive joint time frequency (AJTF) and particle swarm optimization (PSO) algorithms are performed to estimate the phase signals of prominent scatterers. These estimates are used to determine phase nonlinearity. However, such estimation algorithms undergo multi-dimensional optimizations, which yield significant increases in computational complexity for ISAR frame-selections. For more practical use of AJTF- and PSO-based frame-selections, we introduce a method to improve the computational efficiencies of the conventional methods. To this end, in the proposed method, a simple phase unwrapping-and-least square fitting method is used as an alternative to the AJTF and PSO methods. Because PU-LSM can provide reliable phase estimates quickly without complex optimization procedures, the proposed method is an improvement over AJTF- and PSO-based methods, especially in terms of computational efficiency. In simulations using ideal point scatterers, our proposed method performs well, in terms of frame-selection ability and computational efficiency. Furthermore, experimental results using real radar echoes validated the robustness and efficacy our proposed frame-selection method.