用于图像边缘检测的SOC FPGA系统  

SOC FPGA System for Image Edge Detection

作　　者：郝振中 余耀 赵东 HAO Zhenzhong;YU Yao;ZHAO Dong(Nanjing University of Information Science and Technology,Nanjing 210044;Wuxi University,Wuxi 214105)

机构地区：[1]南京信息工程大学,南京210044 [2]无锡学院,无锡214105

出　　处：《计算机与数字工程》2024年第8期2317-2322,共6页Computer & Digital Engineering

摘　　要：图像处理领域中,往往存在着软件处理实时性较差、FPGA硬件功耗大、片上资源使用较多的问题。为此,论文设计了一种资源消耗少、处理速度快的图像边缘检测系统。在FPGA上实现图像采集、图像灰度化、二值化、形态学滤波、优化的Canny边缘检测算法等功能,由硬核处理器实现外设及其IP核的映射,并配置摄像头、DDR3等设备的驱动程序。通过AXI总线协议实现HPS到FPGA的桥接,完成整个SOC FPGA系统的搭建。图像检测结果表明,该图像处理系统资源消耗低、实时性较高,继承了SOC与FPGA的双重优势,为图像处理方法与应用提供了新的发展方向。In the field of image processing,there exists problems of poor real-time performance of software processing,high power consumption of FPGA hardware,and wasted uses of on-chip resources.To this end,this paper provides an image edge detection system with low resource consumption and fast processing speed,realizing image acquisition,image grayscale,binarization,morphological filtering,optimized Canny edge detection algorithm and other functions on FPGA.A hard processing system allows realizing the mapping of peripherals and their IP cores,and configures drivers for cameras,DDR3 and other devices.The bridge from HPS to FPGA is realized through the AXI bus protocol to complete the construction of the entire SOC FPGA system.The image detection results show that the image processing system,which inherits the dual advantages of SOC and FPGA,has a low hardware consumption and high real-time performance.This system provides a promising development direction for image processing methods and applications.

关 键 词：片上系统 边缘检测 硬核处理系统 FPGA 

分 类 号：TP37[自动化与计算机技术—计算机系统结构]