A passive MMW radiometric image feature extraction for security inspection

Abstract

For application of security inspection, the passive millimeter wave (MMW) radiometric transferring model is presented in the indoor experiments. A simulation imaging experiment to the human body with concealed metal objects was done by using 3mm waveband alternating current radiometer. A feature extraction method for millimeter wave radiometric image is also proposed based on morphology and contour method. The results show that the contour processing method can overcome the disadvantage of untrue edge phenomenon in traditional differential edge extraction method and the features extracted of objects are more conformed to the geometrical characteristics of concealed metals.