Research On Ship And Ocean Characteristic Analysis For Ship Detection Using Polarimetric SAR Data

Synthetic aperture radar (SAR) is an important microwave remote sensor, owning the unique imaging capabilities of all weather, all time and other advantages. Polarimetric SAR data can provide a favorable data support for SAR image interpretation, due to its abundant scattering information of targets. Polarimetric SAR data for ship detection plays an important role in many fields, such as maritime transport, fishery monitoring and oil spill detection. The characteristic of ship and ocean is researched on the dual-polarimetric and quad-polarimetric SAR data. Based upon it, these methods of ship detection are proposed in this dissertation.The main work and innovation of this dissertation are as follows:(1) A novel ship detection method using the polarization distance feature is proposed. The different characteristic of ship and ocean is described by polarization distance feature, which effectively separate the ship from ocean. A CFAR ship detection method for polarimetric SAR data is conducted by utilizing constant false alarm rate (CFAR) theory, which obtains a good detection effect.(2) A novel ship detection method based on polarization shannon entropy is proposed. It analyses the different response characteristic of shannon entropy for ship and ocean, which is available to realize ship detection, as well as the improvement of efficiency and precision by using shannon entropy for threshold segmentation.The C band RadarSat-2 dual-polarimetric SAR data and L band UAVSAR quad-polarimetric SAR data are used for experiment. The results demonstrate that the proposed polarization distance feature method and shannon entropy method both have the lower probability of miss detection and higher detection accuracy than other methods, under the same probability of false alarm.The proposed method of this dissertation would be a meaningful progress towards ship detection from polarimetric SAR data.