Cherry Tree Monitoring by a Ground-based SAR System and its Polarimetric Analysis

Abstract

Synthetic aperture radar (SAR) is a well-known technique for airborne or space-borne remote sensing. It can usefully be exploited in a ground-based radar imaging system. We have developed a broadband ground-based polarimetric SAR (GB-SAR) system for environmental studies. Testing results show satisfactory polarimetric performance of developed system. Polarimetric calibration obviously improved the features of the GB-SAR system. Using the developed SAR system, we carried out measurements for a cherry tree in three different conditions, namely in buds before blooming, blooming and voluptuous leaves, respectively. Three-dimensional images were reconstructed from the acquired data. Analyzing the polarimetric images of each measurement, differences among the different polarizations were found. This system provided valuable information about the scattering features of the cherry tree in different periods, which can be detected by broadband polarimetric ground-based SAR measurements. Further, different polarimetric analysis techniques were employed to interpret acquired polarimetric data. The single frequency color images of power density for the covariance matrix are created. Then, an eigenvector based decomposition algorithm is used for identifying the target scattering characteristics with frequency dependency, as well as changes in different situations. The results of polarimetric analysis are presented in this report, which show good agreement with the ground truth.