Through The Wall Radar Imaging Algorithm Of Multi-angle Study

Through-the-wall imaging radar (TWIR), which can use certain through wall imaging algorithm to obtain the image of the scanning region after obstacle by transmitting the electromagnetic wave through the electromagnetic non-transparent dielectric and processing the received echo signal. Therefore, TWIR system will possess a huge perspective in anti-terrorism, disaster relief and other military-civil fields.Multiple-viewpoint through wall imaging not only expands the capability and the value of the TWIR, as the significant application branch of the TWIR system, but also can acquire the building layout image, which is the puzzle for the single-viewpoint through wall imaging. Because of the increment of echo information, effective processing and reasonable use raise a higher claim for the multiple-viewpoint TWIR, bring many new theory and technology problems as well. To solve such problems, the main work is developed as follows:1.The fundamentals of the wideband beamforming for TWIR are presented, and three kinds of multiple-viewpoint radar arrangement are researched. In addition, two evaluation methods of image fusion performance are analyzed, which are verified to be uniformly valid by simulation results.2.For the problems of clutter and multipath effect which through wall radar faced, present a novel algorithm named signal subspace process(SSP). This method can suppress the clutter in radar image via immensely match the data of different channels. Simulation results and experiment results are used to valid the method.3.For the problems of multiple-viewpoint images fusion of dim targets, propose a direct image fusion algorithm based depth first searching. It makes a better performance in fusing images by combining direct threshold target detection and compensating dim target. Experiment real data is used to verify the effection of the method.4.For the issues of target detection in multiple-viewpoint images, study the likelihood ratio modle of target judgement based NP norm, by the statistical perspective. Redouble, fitting and iteration are introduced carefully which are two kinds of solution of the mentioned model. Experiment result gives the performance compare between two model solutions and certain relate conclusion.In this paper, all the researches of these models and methods of multiple-viewpoint through wall imaging are based on a typical background. Two proposed new algorithms are easy to project implementation and have better instruct signification. The details can be constructed the solid foundation for the following study as well.