Processing of Envisat Alternating Polarization Data for Vessel Detection

The alternating polarization (AP)-mode data from Envisat/Advanced Synthetic Aperture Radar (ASAR) can simultaneously generate two well-georegistered images in different polarizations, including VV/HH, HH/HV, and VV/VH combinations. This additional polarimetric information between channels is helpful for enhancing the accuracy of vessel detection. In this letter, different processing methods of AP single-look complex (APS) and precision (APP) data for vessel detection are investigated. For APS data, we propose a dual-polarization whitening fusion filter to combine dual-channel data and present its superiority to other traditional detectors. For APP data, based on the distinct polarimetric properties of each polarization, we utilize different fusion methods for different polarization combinations, including a wavelets-based method for VV/HH and a simple product fusion method for the other two pairs including the cross-polarization channel (HV or VH). Finally, in the experiments, we use real Envisat AP data sets to verify the method, and the results of our proposed processing strategy indicate that it cannot only suppress the speckle and the sea clutter but also improve the vessel detection performance.