Terahertz radar imaging based on block sparse Bayesian learning framework

There is an increasing interest in high-resolution radar imaging of objects, and recent developments of terahertz sensing techniques provide the depiction ability of objects in detail. In this paper, the compressed sensing theory is introduced to terahertz radar imaging. A terahertz radar azimuth-elevation imaging scheme based on block sparse Bayesian learning framework is proposed. By exploiting block sparse structures of the terahertz azimuth-elevation imagery, the reconstruction performance can be improved significantly. Simulation results based on electromagnetic calculation data show that the block sparse Bayesian learning algorithm keeps a better balance between the computation load and the accuracy of the reconstruction signal than the existing algorithms.