FAST 3-D MICROWAVE IMAGING METHOD BASED ON SUBAPERTURE APPROXIMATION

In this paper, the subaperture approximation (SA) method for 3-D microwave imaging is presented based on the sparsity of 3-D image. The idea is that the sparsity information can be extracted from the lower resolution image obtained using the subaperture of the (virtual) array and be used for high-resolution imaging to reduce the imaging region. Thus, a recursion procedure that can significantly reduce the computational cost is established. Compared with the surface-tracing-based method, the SA method can avoid the loss of isolated scatterers. The feasibility is verified by using experimental data. After analysis, the SA method can reduce the computational cost from two aspects: reducing the array element number needed to be processed and the pixels needed to be processed. The computational cost is mainly related to the target characteristics (the sparsity ratio and the topological structure), and decreases with the increase of the sparsity ratio. When the sparsity ratio is larger than 97.6%, the computational cost can be lower than 10% of the 3-D back-projection (BP) method.