The Study And Design Of Real-time Image Processing Platform Based On FPGA

Currently, the information contained in video images is becoming more abundant, the resolution is also rising, meanwhile, the market demand for related products more and more urgent, which requires improving the design requirements and also shortening product R&D cycle, the reconfigurable nature of FPGA’s is to meet such market demand. At the same time in the system design, real-time image processing requirements of the processing speed is relatively high, the traditional serial software design is hard to meet the requirements, in order to meet the requirements of real-time, currently FPGA devices is widely used as an ideal. Therefore, the real-time image processing platform research and design based on FPGA has been much concern.The hardware design platform of my image processing system based on Xilinx Spartan series of low-cost core chip, software using ISE, Modelsim and ChipScope, and other related software. This design, from the system point of view, the use of top-down design principles, using Verilog HDL hardware description language modular design, build FPGA-based real-time image processing platform. Platform consists mainly of the cache control module, video encoding and decoding modules, and image processing module. Image processing module in which use an adaptive image processing method based on the traditional image processing methods. In order to get better real-time image processing effects, the system evaluate the image reference to objective criteria, making different decisions by the image quality, then using different processing algorithms.Finally, testing and simulation the image processing platform, the hardware environment and related algorithms. Testing the hardware environment with loop video experimental, Matlab co-simulation verification algorithm, the results show that the hardware design is reasonable and the algorithm is effective.