Remote human respiration detection using ultra-wideband impulse radar mounted on a linearly flying platform

Abstract

Non-contact vital sign detection using radar is relevant for many applications. In search and rescue missions in disaster-stricken areas, this technology can be used to non-invasively detect live survivors on the ground. However, in a very large disaster area, a fast and effective detection approach is required. This need has suggested radar mounted on a flying platform such as a drone as the most feasible approach. This task is challenging, since human respiration is weak, and the signal recorded is easily affected by disturbances such as noise and movement of the platform. Therefore, in this study, we propose a signal processing technique to deal with this problem. Human respiration signals modulate a hyperbolic pattern recorded by moving radar because of distance projection, leading us to applying sequential image processing steps and hyperbolic pattern reconstruction to extract respiration signals. A Fourier transform is then applied to seek the respiration frequency component. The results of laboratory experiments show that the proposed method can detect human respiration. As an important note, the flying speed of the platform should be determined carefully to cope with slow human respiration.