A Fast Level Set Method for Multimaterial Recovery in Microwave Imaging

作者：Shah Pratik^*; Moghaddam Mahta

来源：IEEE Transactions on Antennas and Propagation, 2018, 66(6): 3017-3026.

DOI：10.1109/TAP.2018.2817278

• Born iterative method (BIM)
• electromagnetic tomography
• gradient methods
• inverse scattering
• level set method
• microwave imaging

Level set-based methods for microwave imaging can produce better results than linearized approaches but they are usually very slow because they require many iterations of a forward solver. In this paper, we have developed an approach, where we need only a small number of forward solver runs, to reconstruct the microwave images accurately. Our approach is a combination of a level set method and a linearized approach and has a regularization term for stability during level set iterations. We provide a detailed analysis of the method and demonstrate its applicability on synthetically generated data for 2-D microwave imaging. Different objects chosen to represent realistic features are considered to evaluate the performance. This approach is extended to multiple materials. The reconstructed images indicate that the method can produce accurate object localization, and shape and size identification. It can also recover the shape very well even for high contrast and heterogeneous structures.

• 出版日期2018-6