文章摘要
宋杨,殷焕松,孙丙宇.基于FPGA的实时图像预处理系统研究[J].安徽建筑大学学报,2024,32():
基于FPGA的实时图像预处理系统研究
Research on FPGA-Based Real-time Image Preprocessing System
投稿时间:2023-05-08  修订日期:2023-06-15
DOI:
中文关键词: FPGA  双边滤波  边缘检测  视频图像处理  图像去噪
英文关键词: FPGA  bilateral filtering  edge detection  video image processing  image denoising
基金项目:
作者单位邮编
宋杨 安徽建筑大学 电子与信息工程学院 230601
殷焕松* 安徽建筑大学 电子与信息工程学院 230601
孙丙宇 安徽建筑大学 电子与信息工程学院 
摘要点击次数: 136
全文下载次数: 49
中文摘要:
      在对图像去噪的实时性要求和对图像边缘保护要求越来越高的背景下,在已有的条件下设计了一套基于FPGA的实时图像预处理系统,在图像处理部分针对图像采集过程中出现的各种随机噪声,在硬件开发平台上基于FPGA实现了双边滤波算法,sobel边缘检测算法,膨胀算法,腐蚀算法的设计,并将处理后的图像在TFT LCD显示屏上进行显示,通过仿真及实验设计,对640×480分辨率图像进行处理,在PC上耗时1.52s,采用本研究所给方案耗时为14.3ms,结果表明该系统既可以达到降噪平衡,同时又保持了边缘的效果,并达到了30帧/秒的640×480分辨率实时显示的图像预处理功能,为后续其他算法的处理创造了良好的条件。
英文摘要:
      Under the background of real-time requirements for image denoising and increasingly high requirements for image edge protection, a set of FPGA-based real-time image preprocessing system is designed under existing conditions, and in the image processing part, for various random noises that appear in the image acquisition process, the design of bilateral filtering algorithm, sobel edge detection algorithm, expansion algorithm, and corrosion algorithm is realized on the hardware development platform based on FPGA, and the processed image is displayed on the TFT LCD display. Through simulation and experimental design, the processing of 640×480 resolution image takes 1.52s on PC, and the scheme given in this paper takes 14.3ms, the results show that the system can achieve noise reduction balance while maintaining the edge effect, and achieve the image preprocessing function of real-time display at 640×480 resolution of 30 frames per second, which creates good conditions for the processing of other algorithms in the future.
查看全文   查看/发表评论  下载PDF阅读器
关闭

分享按钮