Study On Wide-angle ISAR Motion Compensation And Imaging Methods

Because inverse synthetic aperture radar(ISAR)detects target actively,ISAR can work at ultra-long distance in day-and-night under all-weather conditions.Thus,ISAR imaging has become a key to target detection,tracking,classification and recognition applications.Due to the two-dimensional(2D)resolutions of ISAR imaging increase with the rotational angle,wide-angle ISAR(WAISAR)imaging has become an effective and important way to obtain high-resolution ISAR images.However,since the ISAR observing targets are typical-ly uncooperative and the coupling between translational motion(TM)and rotational motion(RM)becomes strong as the rotational angle increasing,accurate motion compensation(Mo-Co)is necessary for high-resolution WAISAR imaging.In addition,the low signal-to-noise ratio(SNR)environment and near distance between radar and target can also degrade the image quality greatly for WAISAR imaging.Aimed at these problems,with the beginning of WAISAR imaging,this dissertation focuses on the problems of accurate Mo Co for ISAR imaging in low SNR environments,WAISAR motion decoupling,estimation and compensa-tion,near-field WAISAR imaging and WAISAR three-dimensional(3D)imaging.The main contents of this dissertation are summarized as follows:1.Accurate Mo Co for ISAR imaging in low SNR environmentsAccurate Mo Co is the key to ISAR imaging,however,the noise would decrease the accuracy of Mo Co,especially in low SNR environments.Aimed at this problem,an accurate ISAR imaging Mo Co is proposed based on adjacent interference-low pass filtering(AI-LPF)in Chapter 2.By taking advantage of the 2D coherence of ISAR echo,the energy of all scat-terers in the target are focused at a small peak.At the same time,the motion parameters are estimated in this focusing process by an iterative way.Theory and simulation results validate that the variances of the estimated TM parameters are close to the corresponding Cramér–Rao low bound(CRLB)and the proposed method has high performance on accu-racy,efficiency and de-noising.2.Jointly estimating TM and RM for WAISAR imagingWAISAR imaging is an effective way to obtain high-resolution target image.However,the coupling between TM and RM becomes strong as the rotational angle increasing,which increases the difficulty of ISAR Mo Co.Aimed at this problem,the motion coupling before and after polar format algorithm(PFA)processing is analyzed and the result shows that PFA can effectively remove most of the coupling between TM and RM even when the rotational parameters are wrong.Based on this finding,a method of jointly estimating TM and RM for WAISAR imaging is proposed.First,the motion coupling is greatly reduced by performing TM compensation(TMC)and PFA with inexact motion parameters,then the residual TM is estimated from the independent azimuth error phase and the TMC is afterwards performed with the estimated residual TM parameters for the data after PFA.Finally,RM parameter is estimated from the range-dependent azimuth high-order error phase.This processing will be repeated for WAISAR imaging and at the same time,motion parameters are also updated iteratively.The iteration wouldn't stop until the image is focused,e.g.the difference of image entropies between the two adjacent iterations is smaller than the threshold or the maximum iteration number is reached.The experiments results validate that the proposed method can effectively overcome the strong-coupling-caused ISAR motion estimation and compensation degrading and achieve high-resolution WAISAR focusing imaging.3.Near-field WAISAR imagingWith the increasing of rotational angle,improving of image resolutions and decreasing of the distance between radar and target,the far-field assumption will be invalid,and this will lead to the distortion and defocusing problems of the traditional far-field-based ISAR imag-ing algorithms.Aim at these problems,two different near-field ISAR imaging algorithms are proposed in the Chapter 4 and 5,respectively.In Chapter 4,with the beginning of invari-ance in azimuth(IIA),the reason of IIA losing for near-field ISAR is analyzed,and an IIA recovery method based on sub-aperture decomposition is afterwards proposed,then by per-forming Stolt interpolation and 2D IFFT,the focusing image without distortion is obtained.Experiments results validate that the proposed near-field ISAR imaging algorithm can solve the defocusing and distortion problems even when the target is moving with TM.In Chapter 5,starting with the spectrum analysis,the 2D wavenumber spectrum analytical expressions of near-field and far-field ISAR are derived respectively by the time-frequency reversion method.The result shows that the phase difference between them is an indepen-dent term.Base on this finding,an extended PFA(EPFA)is proposed for near-field WAISAR imaging.First,the near-field ISAR is equivalent to the far-field ISAR by spectrum correc-tion,the RM correction is afterwards performed by the PFA,finally,the focused image without distortion is obtained after 2D IFFT.The proposed EPFA can achieve 360° rotation-al angle near-field ISAR imaging.Compared with Back Projection Algorithm(BPA),they have similar accuracy but the proposed EPFA is much efficient than BPA.4.WAISAR 3D imaging based on structured sparse optimizationThe wavenumber spectrum of azimuth WAISAR is 3D when the elevation angle is not fixed at 0,which leads to the height-dependent defocusing problem of 2D ISAR imaging which is hard to solve.However,on the other hand,this means the WAISAR has the 3D resolvable ability.According to this,a WAISAR 3D imaging method based on structured sparse opti-mization is proposed in Chapter 6.First,the target is transformed into 3D turntable model by TMC,then the nonuniform FFT(NUFFT)is applied to fast correct the RM,2D match filtering is afterwards performed to obtain the 2D slice images at different heights to form the initial 3D image cubic with high side-lobes.Finally,by constructing structured sparse optimization and solving it using the alternative direction method of multipliers(ADMM),the enhance 3D image can be obtained,in which the side-lobes are much mitigated,and the resolutions are improved.The experimental results validate that the proposed WAISAR 3D imaging method can obtain high-resolution 3D image of target with much less samples.