Fast estimation of high-order motion parameters for real-time ISAR imaging

摘要

The inverse synthetic aperture radar (ISAR) technique is an important tool for target recognition and classification; thus, the high-quality and real-time performance are two essential indicators for ISAR imaging. Based on the classical range-Doppler principle, the motion compensation is the prerequisite step for the subsequent imaging processing. Due to the non-cooperative characteristic of the target, the unknown moving parameters are required to be well estimated. Generally, the translational motion of the target can be accurately described by two-order dynamic parameters. However, the most parameter estimation methods can only estimate the one-order parameter, while the common high-order estimation methods require priori knowledge and are complex to implement. Aiming at this issue, the authors propose a high-order symmetric accumulated cross-correlation method to realise the rapid and accurate estimation of the motion parameters with no requirement of priori knowledge. It takes the advantage of the symmetric accumulation manner to offset the phase errors and optimise the computational complexity simultaneously, and then formulates the estimation to solve the least-square problem. Experimental results verify that the proposed method shows distinct advantages on achieving the high-accuracy and low-complexity parameter estimation, which is highly conductive to realise the high-quality motion compensation for real-time ISAR imaging.