Research Of High-Speed Image Acquisition And Processing System Based On FPGA

With the rapid development of information technology, the importance of image acquisition and processing system is increasingly prominent. Today, image acquisition and processing system is widely used in these fields such as military, industrial, medical, as a result, the system with the characteristics of high stability, strong ability to deal with the data and low cost has become the focus of research. The processing relying on software is unable to meet the requirement of real-time and ASIC of image processing has the problem of long-time development, so using FPGA to realize these fuctions. FPGA,high flexibility that can can design logic circuit according to demand, has the advantage of parallel processing greatly improve the effect of real-time to realize the effect of the processing.This paper was mainly focused on the high-speed image acquisit ion and processing system of FPGA. In aspect of edge detection al gorithm, mathematical model of edge detection and the theory and characteristics of edge detection algorithm was studied in this paper. For the result of analyzing the treatment effect and algorithm com plexity of the algorithm and the reason that FPGA is suitable for a nd the application of the simple algorithm and with numerous data, Sobel operator and Kirsch operator were finally chosen as the maj or algorithm of the system. Hardware design mainly included the se lection of components. The whole system includes FPGA minimum system and peripheral circuit. Method of camera configuration inclu ded the design and simulation of SCCB control module, VGA displ ay controller and the SDRAM control module. In the implementatio n of the algorithm, color space conversion, Sobel operator and Kirs ch operator were analyzed. Adaptive threshold Kirsch operator was designed based on Kirsch operator. Morphologic closing operation wa s used for further processing. The algorithms mentioned above were all realized by FPGA. According to the image processing results, using Sobel algorithm for image processing generally can get relativ ely complete image edge. Compared with image processed by Sobel operator, edge details of the image was increasing by using Kirsch operator’s eight direction detection. Adaptive Kirsch algorithm can be detected more edge detail of the low brightness and high bright ness region than Sobel algorithm and Kirsch.The whole system was designed by using Verilog language under quartus II software environment and was simulated in modelsim.The results shows that this system can well realize algorithms of image high-speed data acquisition and image processing, and image acquisition of 25 frames, can display the results of image processing, have higher practical application value. The analysis and experimental results show that the system which can better achieve high-speed image acquisition and image processing and can collect the image with the speed of 25 frames per second to show the real-time image processing results have high practical application value.