Feature level fusion of SAR and optical image and MPSoC implementation

SAR image and optical image are heterogeneous images, so the integration of these two types of images can take full advantage of their complementary characteristics to obtain more description of the information. For this application, this paper gives one SAR and optical image feature fusion algorithm and presents a multiprocessor SoC system based on this algorithm. The system uses a master-slave/pipeline structure to solve the problems of communication and synchronization among processors and implement the real time processing for the feature fusion system of SAR and optical image. Mentioned multiprocessor SoC is implemented on a single FPGA chip to achieve the miniaturized design of the feature fusion. This can be used in military and aerospace field.