Ballistic target discrimination based on polarimetric entropy

Polarimetric entropy derived from the Cloude-Pottier decomposition is widely applied to fully polarimetric SAR images for target unsupervised discrimination and classification. In this paper, an algorithm for discriminating ballistic targets based on polarimetric entropy is studied. Using the fully polarimetric narrowband target echoes, polarimetric entropy is extracted to represent the randomness of target scattering characteristic in temporal domain. Based on the different polarimetric entropy values of warhead and decoy, the discrimination method is put forward. Lastly, simulations are performed using the measured data of a scaled target, whose results prove the effectiveness of this algorithm.