LAUSR

Through-the-wall imaging (TWI) of human vital signs by bioradar is a hot research topic in recent years. Unknown wall parameters (mainly thickness and dielectric constant) are huge challenges for TWI. Ambiguities in wall parameters will degrade the image focusing quality, lower signal-to-noiseclutter ratio (SNCR) of vital signs, cause vital signs to be imaged away from their true positions and blur the close vital signs from multiple humans caused by the imaging resolution declination. A through-thewall propagation model of vital signs for multiple-input and multiple-output (MIMO) bioradar is first built to analyze the influence of wall on imaging. In order to obtain focused image of vital signs quickly, an imaging model and a novel autofocusing imaging method of vital signs are proposed in this paper. Since vital signs of human are weak and sensitive to interferences, the SNCR-enhanced imagery of vital signs after change detection (CD) is applied to evaluate the focusing quality of image. Reflections of wall in the stationary targets imaging result are line structure approximately, so Hough transform is used to extract the positions of the front edge and rear edge of wall automatically. Propagation time in the wall of electromagnetic waves is estimated and used to build the constraint relationship of wall parameters. The number of unknown parameters is reduced to only one and the efficiency of autofocusing imaging improves. Several cases, including the case of single human, multiple human objects close to each other and the case of non-human objects, are simulated. The magnetic resonance imaging (MRI) image of human chest is put into simulation scene. And then the simulation data of human vital signs are calculated by the finite-difference time-domain (FDTD) method. The results show that the proposed method can effectively estimate the wall parameters and improve the focusing performance of human vital signs. And also the kurtosis of image can be used as a feature to efficiently decide the human vital signs are existed or not. Thus the SNCR of vital signs and resolution of imaging are improved, which are beneficial for detection of vital signs. The position errors of human vital signs are also corrected.