Research And Implementaion Of Image Processing Algorithm Based On FPGA

In recent years, with the rapid development of computer science and technology,the requirements for real-time image processing are increasing, such as medical image analysis, remote sensing image processing, robot vision, video surveillance, video compression and other fields, image resolution requirements are gradually increased, software-based image processing has been unable to meet the needs of real-time. Today, FPGA with its programmable, reconfigurable, algorithm processing speed of hardware and other characteristics, is ideal for image processing system. At present, the use of FPGA in image processing applications are increasingly attracting people's attention, and have a broader development and application prospect, FPGA-based image processing has become a hot area of research at home and abroad.This paper takes FPGA as platform, designs hardware for median filtering algorithm, fast median filtering algorithm, average filtering algorithm in the image preprocessing, and improves the existing fast median filtering algorithm, the efficiency and effect of algorithm has been greatly improved. The main topic related to the image segmentation algorithm consists of edge detection algorithm and the method of geodesic active contour model. Based on the parallel lines character of FPGA, the paper respectively conduct FPGA design and implementation for Sobel and Prewitt edge detection operator. And compares from both time and result with software. In addition, the article also designs the overall circuit hardware of complex image segmentation model, and simulates the key feature calculation module of the system core.The design and implementation is synthesis and simulation experiment done in Altera's development environment called Quartusii and third-party simulation software named Mdoelsim. The results show that the image processing based on FPGA can obtain good results, and processing speed is much higher than software-based processing speed, can achieve real-time image processing requirements. The subject did a useful attempt exploration to the design of FPGA-based image processing algorithms, and has a positive significance and far-reaching impact for the future of image processing system.