摘要
在对图像去噪的实时性与质量要求日益提高的背景下,该设计根据双边滤波算法的原理,将双边滤波空域核与值域核部分计算方法进行优化,设计了基于FPGA的图像双边滤波器处理系统。该系统利用上位机对FPGA端发送图像数据,FPGA将接收的图像数据经预处理之后对图像进行双边滤波处理,最后将处理完的图像数据按照视频格式采用HDMI接口输出到显示屏上。在对双边滤波模块接口设计时引入AXI总线,使用纯Verilog语言及参数化编程,模块可移植性强。从实验结果可以看出:系统对800×600分辨率噪声图像进行处理,耗时13 ms,处理过后图像的PSNR提高了3.75,SSIM提高了0.22,功耗仅1.3 W。
Under the background of the increasing demand for real-time and quality of image denoising,the design optimizes the calculation method of bilateral filtering domain kernel and range kernel according to the principle of bilateral filtering algorithm,and an image bilateral filter processing system based on FPGA was designed.The system used the host computer to send image data to FPGA,FPGA received the image data after the pre-processing of bilateral filtering image,finally processed image data according to the video format using HDMI interface was output to the display.AXI bus was introduced in the interface design of bilateral filter module,and pure Verilog language and parametric programming were used to make the module highly portable.It can be seen from the experimental results that it takes 13 ms for the system to process noise images with 800×600 resolution.After processing,the PSNR and SSIM of the image are improved by 3.75 and 0.22 respectively,and the power consumption is only 1.3 W.
作者
严飞
陈佳宇
张华西
刘银萍
刘佳
YAN Fei;CHEN Jia-yu;ZHANG Hua-xi;LIU Yin-ping;LIU Jia(Automation College,Nanjing University of Information Science&Technology,Nanjing 210044,China;College of Atmospheric Physics,Nanjing University of Information Science&Technology,NanJing 210044,China;Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology,Nanjing,210044,China)
出处
《仪表技术与传感器》
CSCD
北大核心
2022年第6期99-104,共6页
Instrument Technique and Sensor
基金
江苏省产业前瞻与关键核心技术重点项目(BE2020006-2)
国家自然科学基金项目(61605083)。
