Based On Ct Imaging Synthetic Aperture Radar Vector Entropy Imaging Studies

Synthetic aperture radar (SAR) has been used widely in more and more fields, especially in automatic target recognition systems. Researches on 2-D and 3-D SAR image reconstruction have important theoretical value and applications.CT technology is widely used in industrial harmless-examination, geography explores and so on. Based on CT technology, a new SAR reconstruction algorithm has been presented, which involves a tomographic model of SAR observation and a Feature-Enhanced Vector-Entropy regularization approach to reconstruct SAR images from projections. The performance of the method by using synthetic and real SAR scenes has been demonstrated.A new SAR 3-D model of tomography has been developed by proving that the possessed return signal can be shown to be a certain range of 3-D Fourier transform of reflectivity function. This 3-D tomography provides a new way for understanding the 3-D nature of reflectivity function.In place of regularization method which is used in real-valued image reconstruction , Vector-Entropy regularization method has been developed which effectively deals with the complex-valued, random-phased nature of the SAR target reflectivity and the nonlinearity in optimization.Extraction of features from SAR images is important for automatic target recognition system. Our algorithm is recognition-oriented by incorporating some prior information regarding the nature of the features of interest during the reconstruction and helps provide reconstructed images with better enhanced target features.The proposed algorithm has been tested by both synthetic SAR scene and data from MIT Lincoln Laboratory ADTS dataset and then shown some advantages over the traditional SAR technique on the reconstructed quality and computation time.