The Method Of Interferometric Three-dimensional ISAR Imaging

Since the concept of inverse synthetic aperture radar(ISAR)was put forward,it has been widely concerned and applied in military and civilian fields such as space target detection,battlefield reconnaissance,anti-missile early warning,civil aviation control and so on,with its unique advantages.Interferometric inverse synthetic aperture radar(InISAR)can obtain the three-dimensional images of non-cooperative targets by combining ISAR imaging with interferometric technology.Compared with ISAR imaging,the target information obtained by interferometry ISAR is completer and more abundant,and it can directly reflect the threedimensional structure of the target,and is less affected by the target movement and attitude adjustment,which has important value and significance for improving the ability of target feature extraction and recognition of radar.Therefore,this paper focuses on the key issues of image registration and squint problem in interferometric ISAR imaging,and proposes an image registration method based on residual translation component compensation and a squint InISAR image registration and three-dimensional imaging method based on particle swarm optimization(PSO)algorithm.The main contents are as follows: 1.The principles of ISAR imaging and InISAR three-dimensional imaging are studied.Based on the ISAR imaging echo model,two kinds of range compression methods,matched filtering and Dechirp,are discussed,and the physical significance and common methods of translational motion compensation are clarified.Then,based on the "L-shape" three-antenna InISAR imaging system,the interference echo signal model is constructed,the relationship between interference phase and scatterer coordinates is deduced,and the three-dimensional imaging process of InISAR is summarized.2.The image registration method of InISAR three-dimensional imaging is studied.For the mismatch of different ISAR images on the same baseline,based on the three-dimensional angular motion model of the target,the causes of image mismatch and its influence on interferometric imaging are analyzed in depth,and the analytical expression of image mismatch is deduced.Then,the application scope and shortcomings of two traditional image registration algorithms are discussed in depth.On this basis,a new image registration method based on residual translation component compensation is proposed.The residual translation component caused by the unified translation compensation based on the echo of the central antenna is discussed in depth,and the influence of the residual translation component on the image mismatch is analyzed.Based on this,the expression of residual translational component is deduced,and the compensation factor of translational component is constructed to compensate the echo of auxiliary receiving antenna,which realizes the accurate registration of different ISAR images.Finally,the effectiveness of the proposed algorithm is verified by simulation results.3.The method of InISAR image registration and three-dimensional imaging under squint mode is studied.For the problem that the interferometric phase ambiguity and the increase of radar angle measurement error in squint scene makes many traditional image registration methods no longer applicable,a new image registration method based on PSO algorithm is proposed.Because of the different position distribution of interferometric antenna,the echoes of different antennas have different wave path.This method transforms the image registration problem into an unconstrained optimization problem for the estimation of target angle parameters by deducing the relationship between the variation of wave path and the angular velocity of the target.With the correlation of two ISAR images as the objective function,the PSO algorithm is used to obtain the optimal estimation of angular velocity and construct the phase compensation factor,then the echo is compensated to achieve accurate registration of ISAR images.Secondly,for the problems of distortion and deflection of threedimensional image caused by squint additional phase,the causes of squint additional phase and its influence on three-dimensional imaging are analyzed in depth.Then a squint additional phase compensation method is proposed.By constructing additional phase compensation factor for squint,the interference phase is corrected,and the real coordinates of the target along the baseline direction are calculated.Finally,the rotating matrix of radar coordinate system and radar line-of-sight coordinate system is used to calibrate the target ordinate,so that the accurate three-dimensional structure of the target can be reconstructed.Finally,the effectiveness of the proposed algorithm is verified by simulation results.