摘要
针对MATLAB对数字图像实时处理速度慢的问题,提出采用硬件并行计算的思想,通过FPGA硬件实现Sobel边缘检测算法。首先对OV5640摄像头采集到的实时图像进行彩色图像转灰度处理,再对输入的灰度图像进行高斯滤波处理,接着对滤波后的灰度图像进行二值化,再使用Sobel算法对二值化图像实现图像边缘检测,传送至SDRAM存储器进行缓存,最后通过VGA实时显示图像。通过图像边缘检测的FPGA硬件处理和MATLAB软件处理的效果对比,结果表明FPGA硬件实现的边缘检测算法效果更佳。
To address the problem of slow real-time processing of digital image by MATLAB, the idea of hardware parallel computing is proposed to implement Sobel edge detection algorithm by FPGA hardware. First,the colored real-time image collected by OV5640 camera is transferred into gray, which then processed with Gaussian filter, and binarized after that. The binarized images go for image edge detection with Sobel algorithm,sent to SDRAM memory for caching. Finally, the images display in real-time through the VGA. By comparing the effect of FPGA hardware processing and MATLAB software processing, the result shows that the edge detection algorithm implemented by FPGA hardware is better.
作者
李芳
沈培
温娜
刘慧迪
辛锋
LI Fang;SHEN Pei;WEN Na;LIU Hui-di;XIN Feng(School of Mechanical and Electronic Engineering,Pingxiang University,Pingxiang Jiangxi 337000,China)
出处
《萍乡学院学报》
2022年第3期83-86,共4页
Journal of Pingxiang University
基金
萍乡学院青年科学基金(2019D0210)
江西省高等学校大学生创新创业训练计划项目(S202010895020)
江西省教育厅科学技术研究项目(GJJ191154)。
