Threat object classification with a close range polarimetric imaging system by means of H-alpha decomposition

In this paper, an approach to differentiate between various dielectric threat objects in security applications is investigated. The scattering information in form of the Sinclair matrix of relevant scenarios is gained from a fully polarimetric, monostatic synthetic aperture radar (SAR). A possible polarimetric calibration procedure is presented. The radar data are processed with the H - α decomposition algorithm. The H - α scattering characteristics of threat objects are analyzed in terms of a weighted averaging. It is shown that an object classification is possible even for threat objects conceiled under thick layers of clothing. Measurement results are presented to illustrate the topic.