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

Nowadays,with the rapid development of science and technology,people have a variety of ways to obtain information.For the transmission of information,the most intuitive is the transmission of image information.Image is the most direct and fastest way for people to perceive things.It is an important information transmission medium.Among them,edge information is one of the more important information of image information.It plays a very important role in extracting some features.At the same time,it is also an information that people are more concerned about.With the development of digital image technology,the transmission of image data is becoming larger and larger.The problem of slow transmission is a common problem.At the same time,because it is not particularly convenient for computers to use software for calculation,using FPGA to design a platform will be more suitable for image edge calculation.Due to the parallelism of FPGA,At the same time,it can design a separate ASIC for design,which has good portability and greatly facilitates image processing.By designing a real-time image edge detection system based on FPGA,this thesis solves the shortcomings of insufficient real-time performance and long time-consuming of traditional image processing when using software in traditional computer.At the same time,by improving the traditional edge detection algorithm,the 8-direction convolution method is used for convolution,and the adaptive double threshold method is used for edge detection.This thesis mainly works through four modules,mainly including camera acquisition part,storage part,algorithm processing part and so on.The edge detection of the image is realized by using the FPGA chip.In this thesis,the ov5640 camera is used to collect the image,the collected data information is stored in the SDRAM memory,the algorithm module is realized in the FPGA chip,the Sobel operator is used to detect the edge of the image,and finally the edge information is displayed by VGA.This design mainly uses FPGA to design a hardware platform,and realizes the algorithm in FPGA chip.Using FPGA parallel processing and pipeline design,we can speed up the operation speed of image processing.However,the current CPU processing is calculated in serial mode,which is not as good as FPGA hardware platform.Therefore,this thesis selects FPGA as the main control chip for design,and carries out edge detection through Sobel operator with adaptive threshold,At the same time of accelerating image processing,the quality of edge detection is improved by using the method of adaptive threshold.Finally,it is displayed on the display by controlling VGA.The main innovation of this thesis is to realize edge detection on FPGA hardware platform through 8-direction double threshold Sobel operator and adaptive threshold method,which is finally displayed.This thesis mainly aims at the design of ov5640 camera control,VGA display control and SDRAM controller.At the same time,it needs to realize IIC protocol,median filter and edge detection algorithm on FPGA chip.Finally,through a series of designs,the image edge is processed and displayed in the display.