Compressed sensing pulse-echo mode THz tomography

This paper proposes a pulse-echo mode terahertz tomography technique for endoscopy applications. It can be shown that the sensing matrix of the pulse-echo mode acquisition has sufficiently small mutual coherence that allows the use of recently proposed compressive sensing theory in signal processing area. More specifically, if the target is sparse, then reconstruction algorithm such as the basis pursuit or greedy method can be applied to reduce the number of measurements and accelerate the acquisition process. Simulation results demonstrate that quite accurate reconstruction can be obtained from very limited number of measurements where the conventional Kirchhoff migration methods fails.