Research On Image Reconstruction Algorithm Of Millimeter Wave Three-dimensional Holographic Imaging

Millimeter wave imaging has been widely applied in domains such as contrabands detecting in public areas, instrument landing system, battle field environment surveillance etc. for its better penetration ability compared with infrared. This thesis introduces millimeter wave 3D holographic imaging system based on optical holography, focuses on researching an image reconstruction algorithm for flat scanning imaging method of millimeter stepped frequency pulse radar, analyses the modeling of Stepped Frequency Pulse Signal (FSPS) and the derivation of holographic radar imaging model and contains one dimensional range profile (1DRP), two dimensional coronal profile (2DCP) with two dimensional cross-section (2DCS) and three-dimensional holographic imaging (3DHI).An interpolation method is derived and used for solving the domain inhomogeneous problem caused by dispersion relations during the transformation from wave number domain to space domain in 2DCS and 3DHI algorithm. The result by reconstructing data in imaging experiment which was carried out in laboratory using flat scanning imaging system for simulated target (metal plane model reduced in scale) has proved the effectiveness of the method mentioned in this article.Another holographic imaging algorithm based on NUFFT is also introduced and applied in the procedure of 2DCS and 3DHI algorithm, which is derived in detail, improved and realized by program as well. Both the imaging accuracy and time for reconstruction are significantly improved in comparing with the results of traditional imaging algorithm in wave number domain.Finally, a holographic algorithm based on optimized K-space data extrapolation's match filtering method is mentioned in order to resolve the problem that the imaging system might have to lose or abandon some data which is caused by complicated outdoor environment, hardware system issues or unreliable sampling. The algorithm procedure is derived and verified by program simulation. Its correctness is proved with program simulation. The experimental data is reconstructed by using optimized algorithm and its outcome is compared and analyzed with the one of traditional wave number algorithm, which validates the practicability of the algorithm.