Scattering Characters Analysis And Detection Algorithm Of The Target For FOPEN SAR

FOPEN SAR has shown the ability to penetrate the tree canopy and detect the large target concealed by tree foliage, hence, there has been considerable interest in the army's application, and at present, the scattering characters of the target positioned upon ground is the important and challenging problem in this area. In this thesis, composite excitation method of FDTD algorithm is studied, after modeling the truck, tank, tree and other complex object with minor details, the scattering characters of these targets and trunk are analyzed in depth. Some distinct characteristics between the tree clutters and the scattering wave from target are extracted, furthermore, the detection algorithm of FOPEN SAR is developed, which is applied to detect the target in the data come from our actual experiments.Firstly, the basic principle of FDTD algorithm is described in detail, the composite excitation method, which is suitable for the scattering analysis of target positioned upon ground, is developed. This method takes two steps: first, the field, which is used to excite the target in the following step, must be achieved by studied the scattering characters of the ground; second, the field achieved above is used to excite the target, the wide-band scattering characters of ground target are acquired by a complex computing procedure. There are three key problems on applying this method to the scattering analysis of ground target: 1) to construct the FDTD iterative equation used in dispersive media; 2) to applied GPML (bound condition) to absorb the lossy wave in dispersive media; 3) to construct the principle of the near to far field transformation technology.Secondly, the method, used to model complex target and tree trunk in FDTD algorithm, is studied. As for complex target modeling, a new method of the complex object modeling for FDTD electromagnetic simulation using the Microsoft DirectX is proposed, and the implementation is detailed. The proposed method combines the technology of the computer graphics and the computational electromagnetics, the procedure uses the DirectX to load the model from a CAD file and get the scale, materials, mesh (an object made up of a set of polygonal faces) of the object directly. Then a series of tracing rays to the model from two or more sides are shot, by analyzing the points of intersections, whether or not the tentative cell lying inside the model is determined, after recording the result along with the electromagnetic parameter got from the materials, the FDTD geometry- electromagnetic model is given. As for tree modeling, according to the results of the standard literature and our experiments, the tree may be simplified to the trunk without lost of the analyzingprecision in this frequency band, furthermore, the trunk is regarded as a lossy dielectric cylinder with reason.Thirdly, the scattering characteristic of the ground for several typical electromagnetic parameters is analyzed, the exiting field is acquired on every case. Furthermore, the scattering data of the DONGFENG truck, M60 tank (TAIWAN), M2A2 (US), Brdm (US), F16 (US), trunk (for several different electromagnetic parameters) and other complex object is obtained as the result of the composite excitation method; from these scattering data, scattering characters of these targets are derived, some SAR image is given as the result.Finally, based on the contrast between the scattering wave of above targets and the tree cluster, detectable electromagnetic characters are derived from the space domain, the frequency domain and the polarization domain. According to these characters, three targets detect methods for SAR image are developed. Firstly, based on the electromagnetic character from the space domain, multi-aperture SAR detection algorithm by using change detection is developed. Secondly, based on the electromagnetic character from the frequency domain, SAR target detection algorithm by using split spectral analysis is developed. The above methods are proved out by the test based on the experiment data of UHF/VHF UWB track SAR system in the NUTD.