A material-intelligent method of terahertz radiation characteristics of target model

Passive terahertz detection technology is used to obtain the brightness temperature image of the target, but cannot directly get target’s physical temperature and material emissivity. In this work, a material-intelligent method is proposed to estimate physical temperature and material emissivity of the target within the terahertz brightness temperature images. Based on the supervised learning algorithm, the multi-task loss function is employed to guide the intelligent algorithm to identify physical temperature and material emissivity of the target in the background. Finally, the simulation results demonstrate that this method is a novel approach to obtain the physical temperature and the emissivity of the material with the brightness temperature images.