Detection of targets beneath foliage using aspect-angle variation of the polarimetric SAR response

This paper summarizes a recent foliage-penetration experiment designed to develop polarimetric SAR algorithms for discriminating between man-made objects and trees. The analytical approach is to exploit differences in the manner in which the polarimetric response of targets differs from that of tree trunks as the aspect angle changes, with the aspect-angle change induced either by flying the SAR through a circle or by building a time sequence of images through sub-aperture processing of data from a linear pass. Our intent is to quantify the polarimetric variation through use of the Cameron decomposition and its associated metric, although other approaches will be investigated as well. This investigation uses a Northrup Grumman Multiband SAR (MBSAR) with the ability to collect sub-meter resolution, polarimetric data, simultaneously in both the L and UHF bands. In June 2013, a field experiment was conducted over a mixed forest in Southern Maryland where a number of canonical targets and vehicles were deployed under the foliage canopy. An array of polarimetric calibration targets was also deployed nearby in an open field. An overview of the experiment is presented in this paper, and the results of an initial analysis of the aspect-angle variation in these data will be presented at the 2014 IEEE National Radar Conference.