Research On The Key Technology Of See-through-building Detection

By the ultra-wide-band microwave signal, see-through-building detection(STBD) is utilized to obtain the wall layout of building and the position information of targets in building such as human, which plays an important role in the areas of anti-terrism, investigation and arrest. The critical point of STBD is to implement the detection for the building layout, stationary target in building, stationary and moving human. Beginning from these four points, the thesis has researched on the techonology problems caused by the wall penetration including the compensation of the electromagnetic propagation distortion, the suppression of the multipath clutter in the enclosed space, the robust tracking of the high mobile and randomly variable moving human, the robust detection of the weak life-sign. The main contributions of the thesis are given as follows:1、Based on multi-channel and multi-view two-stage image fusion, an imaging algorithm of building layout is presented to deal with the problems of the cavities and burrs in wall images and the clutter spots caused by the phase distortion of unknown wall penetration. The algorithm implements the high-quality imaging for the building wall layout, which is better than the existing imaging algorithm with the single-stage multi-view image fusion.2、Based on the image sequences formed by multiple assumed wall parameters and the image focusing with the identical time-delays seperately, two wall compensation algorithms are proposed to address the problems of defocusing and displacement of target image derived respectively from the homogeneous and inhomogeneous unknown wall penetration, which realizes the high-resolution and high-precision imaging for the stationary targets in building.3、Due to the invalidity of the traditional cross-correlation back-projection(CBP) imaging algorithm in the low signal-to-noise ratio(SNR), an improved CBP imaging algorithm is presented to improve the imaging quality in the low SNR and image resolution. Furthermore, we propose a combination CBP imaging algorithm to increase the image SNR by the array element combination and coherent accumulation.4、An adaptive exponent-weighted multiplication approach is presented to fuse multiple single-channel target images. Through applying the image contrast factor to adjust the exponent weights adaptively, the approach deals with the problem of target miss-detection caused by image intensity difference of multiple targets in these single-channel images and improve the signal-to-clutter and noise ratio(SCNR).5、Based on the difference between the stationary human and the other objects, a joint detection approach in the life-sign domain and image domain is designed to implement the detection and localization of the weak human target and the classification of human target and other stationary target. Therefore, the detection ability for the human target is improved in the condition of the low SCNR.6、Including the cores of one-dimension range-delay tracking and two-dimension position tracking and supplemented by the inversion compensation for the localization error caused by wall penetration, a cascading tracking algorithm for the moving human is presented to deal with the problems of multipath interference, random miss-detections and localization error for multiple high-mobility human targets in the enclosed building space, which implements the consecutive, stable and accurate human tracking.Except the improved and combination CBP imaging algorithms validated by the simulation data, the aforementioned algorithms have been proved by through-wall-radar experiment data.