Ground clutter signal is a kind of unwanted echoes in target detection radar system, which are normally reflected by ground surface, ground-based objects, and obstacles. It can be collected and characterized in polar coordinates in terms of range and azimuth. By using polarimetric based algorithms, including them single channel detector, the span detector, the power maximization synthesis (PMS) detector, the identity likelihood-ratio-test (ILRT), the polarimetric whitening filter (PWF), and the optimal polarimetric detector (OPD), the target detection radar system can distinguish the characteristics and diversity of multiple targets.
In this thesis, a polarimetric radar simulator to generate multi-channel polarimetric signals with specific statistical characteristics has been developed and validated in the simulation. In the measurement, a new noise-based equalization for all polarimetric radar channels has been proposed and tested, which improves the reliability and accuracy of the polarimetric information. After noised-based calibration and model-based decomposition of the polarization covariance matrix, with regenerated measurement slow-time data, a variety of targets in heavy clutter with signal power comparable to the target are detected by polarimetric algorithms in the environment of strong clutters. Starting from numerical simulation and comparison of different detector algorithms, this work has validated the feasibility and accuracy of each detector in realistic scenario. The measurement result agrees with the simulation result that with the use of radar polarimetric information as a priori knowledge, target detection can be improved
by polarimetric detectors.