In this paper, a high-performance SoPC-based design of imaging system is presented and its custom video processing is based on reconfigurable computing, which provides both flexibility and performance. Besides hardware framework, its related design tools and methodology is elaborated, from high-level video algorithm exploration and model analysis to low-level real-time hardware implementation attached to an existing embedded platform. In addition, this paper introduces a new simulation and debug way. As a case of study, a real-time video stabilization algorithm is implemented to show the design flow and merits of this platform. In the end of the paper, we compare the design methodology of the past and present to show that new design tools and methods are key to implementing video processing tasks based on FPGA.