New way for concealed object detection using passive THz images without their viewing

Abstract

We developed new real-time algorithm, based on the correlation function, for concealed object detection using computer processing of the passive THz images without their viewing. This algorithm allows us to make a conclusion about presence of forbidden objects on the human body. To increase the THz image quality we propose a new algorithm in comparison with algorithms which have developed by us early. It allows to increase a temperature resolution of the passive THz camera at least 20 times. This approach is based on a correlation function application for computer processing of the raw THz image. The correlation computing occurs between characteristics of the raw THz image, produced by the passive THz camera, and characteristics of a standard image corresponding to one of the detecting objects (knife, gun,…). The standard image moves in two directions along a image under analysis. As a result, 2 D correlation function is obtained. Multiplying this function by color number belonging to a grey scale, we restore the image under the analysis. This allows to suppress a noise on a new image. This algorithm is very convenient for using and has a high performance. Developed approach opens also new type of algorithms for the passive THz image quality enhancing.