Research And Design Of Real-time Digital Image Processing System Based On DSP+FPGA

With the rapid development of computer technology and network technology,digital image processing is more and more widely applied to the military and the civilian area.The high-speed development of microelectronics technology demonstrates more advantages of hardware-based image processing.In this thesis,an image-processing system is studied and designed using the DSP+FPGA architecture.This architecture plays both superiority of DSP and FPGA,and improves the system's performance,shortens design cycles and cost savings.Firstly,the thesis introduces current state of the image-processing system at home and abroad,and compares different kinds of projects.Considering the demands of video image processing,DSP and FPGA are selected as the processors of hardware platform,and the main hardware circuit is designed.DSP,as the main processor,is responsible for image data compression and the processing of data transmitted to the computer's work,meanwhile FPGA,as a co-processor,completes the video image acquisition and preprocessing.Secondly,several common image pre-processing algorithms(including image filtering and edge detection processing methods)and image compression algorithms are expounded.The improved median filtering algorithm and JPEG image compression algorithm are simulated and analyzed,which is the theoretical basis for the realization of the system function.Finally,the median filtering algorithm based on 4 * 3 window template is carried out on the FPGA,which verifies the feasibility of the algorithm.The JPEG image compression algorithm is achieved on the DSP and the Sobel edge detection algorithm and Canny edge detection algorithm are achieved on the FPGA,which improves the function of the entire system.