Microwave Detection of Respiration Rate of a Living Human Hidden Behind an Inhomogeneous Optically Opaque Medium

Different clutter removal algorithms for human life detection are studied and implemented on the synthetic data considering different realistic investigation domains. Suitability of the subspace projection method is explored analytically for inhomogeneous investigation domains. Here, the occurrence of inhomogeneity in the investigation domains is considered along the direction of scan. The algorithms are also implemented on the experimental data collected from an inhomogeneous investigation domain, that has a high probability of occurrence in reality. It is observed that singular value decomposition, one of the subspace projection based clutter suppression techniques is suitable for removing the clutters in all kinds of inhomogeneous investigation domains when compared to the mean subtraction and time gating based clutter removal techniques. Doppler frequency induced due to the thoracic movement of a human being in the through-the-wall case is also detected with a priori information availability.

Automated summary

Different clutter removal algorithms were studied and implemented for human life detection, with focus on the suitability of subspace projection methods in inhomogeneous investigation domains. Singular value decomposition was found to be more effective in removing clutter in diverse inhomogeneous investigation domains compared to other techniques.