Design And Implementation Of Real-time Image Edge Detection System Based On FPGA

The continuous development of science and technology makes digital information grow rapidly.In the traditional digital image processing system,huge and complex image datum is difficult to be processed.Digital image information science and other digital information science not only penetrate each other,but also are used by other disciplines.Furthermore,the modern image and video processing algorithm is more and more complex,which requires more and more real-time,making the traditional image processing system based on software platform difficult to meet the real-time requirements.The most important advantage of FPGA for image processing is as follows: it is capable of real-time pipeline operation,with fast processing speed and the highest real-time performance.Image processing technology has always been a key area of social research today.As an important part of image processing,edge detection technology has been widely used in medical image,machine vision,industrial detection and other fields.At the same time,the actual effects of edge detection will also affect the application of post-processing algorithms,such as feature extraction,target recognition,and target segmentation.Compared with the traditional filtering method and the system implementation of the edge detection operator,this paper proposes a digital image processing system based on FPGA,which combines fast median filtering and improved Sobel edge detection operator to perform fast acquisition and real-time video image preprocessing.This paper introduces the hardware platform of digital image processing system based on FPGA as the main control chip and the implementation of image edge detection algorithm.Analyze the research background and significance of real-time image preprocessing system,compare the current classic mainstream edge detection operators with Matlab simulation software,and select the edge detection operators that are suiTable for the system.Utilize Altera's fourth-generation C8-level FPGA chip controller to control 5 million high-definition camera OV5640,SDRAM memory chip and FTF800 display,and write Verilog hardware description language to design image acquisition,storage,display and image processing through Quartus II development software Module,using Modelsim simulation software to verify the correctness of its logic timing.Finally,experiments show that the FPGA-based edge detection processing platform can quickly process image information in a short period of time.At the same time,the improved Sobel detection algorithm can extract more edge information than traditional algorithms.