Three-dimensional Fourier bases-expansion contraction integral equation for inversion (FBE-CIE-I) is developed to solve electromagnetic inverse scattering problems (ISPs). Taking the shape of the scatterer into account, the size of the domain of interest (DoI) and thus the number of low-frequency Fourier bases are flexibly chosen in the proposed method. The accuracy and the stability of the proposed method are verified by the numerical tests on irregularly shaped targets. The results indicate that the accuracy and the convergence of the proposed 3-D FBE-CIE-I is insensitive to the parameters and the noises on the scattered fields.

An accurate and efficient simulation model is developed for the scattering from a 3-D ship on a 2-D sea-like surface. The ship on the sea-like surface is modeled as a ship on a random surface with the impedance boundary condition. A full-wave simulation method is developed to simulate this model by using the self-dual integral equation and the parallel multilevel fast multipole algorithm. The simulation model is compared with the model of a ship on a flat surface and on a perfectly electrically conducting random surface, respectively. The application ranges of these models are concluded. Furthermore, the characteristics of the ship on a sea-like surface are studied by using the developed simulation model.

An efficient numerical approach is presented for calculating backscattering from rough sea surfaces under low grazing incidence in this paper. The sea surfaces are modeled as random surfaces by using the impedance boundary condition. Different from the conventional integral equation, the self-dual integral equation is used to incorporate the random IBC surfaces. In order to deal with the large simulation domain brought by the low grazing angle, the parallel multilevel fast multipole algorithm (MLFMA) is employed to speed up the simulation. Numerical results of low grazing angle backscattering from 2D rough sea surface are presented to illustrate the accuracy and efficiency of the algorithm. The analyzation of the feature of the scattering from the sea is also provided.

The scattering center model had been proved to be able to characterize the scattering fields from an extended target with high precision. Besides radar image interpretation and radar characteristics recognition, the application of scattering center model is extended to the computation of linear glint errors in this paper. The formula of linear glint errors are derived, as well as the relationship of linear glint errors at two orthogonal directions, which are of significance for the study of probability distribution functions of linear glint errors. The analytical results are validated by the results of the traditional method using the data bank of scattered fields.

It has been proved that scattering center modeling for narrow band signals can be achieved through the optimum image matching of time-frequency representation (TFR). In order to obtain TFR, however, the azimuth sampling interval should be quite small for electrically large target, which results in a huge amount of electromagnetic computation. Under the situation of under-sampling scattered waves, aliasing distortions of the Doppler curves of scattering centers will be caused in TFR, which makes it hard to estimate the parameters of corresponding scattering centers from TFR. To deal with this problem, an approach for scattering center modelling from under-sampling scattered waves are presented in this paper. The random under-sampling scattered waves based on the principle of compressive sensing are applied to acquire the non-aliasing Doppler curves of scattering centers in TFR. The feasibility of this method has been validated by the simulation results.

Millimeter-wave (MMW) radar, which is used for road feature recognition, has performance that is superior to optical cameras in terms of robustness in different weather and lighting conditions, as well as providing ranging capabilities. However, the signatures of road features in MMW radar images are quite different from that of optical images, and even physically continuous features, such as road edges, will be presented as a set of bright points or spots distributed along the roadside. Therefore, discrimination of the radar features is of paramount importance in automotive imaging systems. To tackle this problem, an approach called the stripe Hough transform (HT) is introduced in this paper, allowing enhanced extraction of the geometry of the road path. The performance of the approach is demonstrated by comparison of extracted features from MMW images with the real geometry of the road and with the results of processing by classical HT.