基于SoPC的视频图像去雾设计  

Design of Video Image Defogging Based on SoPC

作　　者：周帅 杨超[1] 印茂伟[1] 李玉林 张雨亭 ZHOU Shuai;YANG Chao;YIN Maowei;LI Yulin;ZHANG Yuting(School of National Defence Science and Technology,Southwest University of Science and Technology,Mianyang 621010,China)

机构地区：[1]西南科技大学国防科技学院,绵阳621010

出　　处：《实验科学与技术》2022年第2期22-27,共6页Experiment Science and Technology

基　　金：全国高等院校计算机基础教育研究会计算机基础教育教学研究项目(2021-AFCEC-477);中国高等教育学会地方大学教育改革发展重大理论与实践问题研究课题(21DFYB08);四川省2021—2023年高等教育人才培养质量和教学改革项目(JG2021-891&JG2021-885);西南科技大学校级教改项目(19xn0014&21szzy01&21GJZXZD03)。

摘　　要：介绍了视频图像快速去雾装置的设计和实现。装置以ZYNQ系列的SoPC(片上可编程系统)芯片作为开发平台,分析基于物理模型的暗通道先验去雾算法并对算法进行优化,设计向下最小值求解器完成单帧图像去雾算法的逻辑实现,提出了基于帧间像素值均值法实现视频去雾算法。通过研究对去雾算法进行优化,增加算法的可移植度同时使用加速技术提升算法运算的速率,实现单帧图像快速去雾处理和视频图像的实时去雾处理,有效地避免了户外成像系统受到雾天的影响,具有十分重要的科学和工程意义。This paper introduces the design and implementation of video image fast defogging device.The device takes the SoPC(system on a programmable chip)chip of zynq series as the development platform to analyze the algorithm of dark channel prior based on physical model and then to optimize it.A downward minimum solver is designed to achieve the implementation of logic of single image haze removal algorithm,and a video defogging algorithm is proposed based on the average value of interframes pixels.By optimizing the defogging algorithm,the portability of the algorithm is increased.At the same time,the speed of algorithm operation is improved by using acceleration technology.The fast defogging process of single image and real-time defogging process of video image are realized,protecting the outdoor imaging system from fog effectively,which is of very important scientific and engineering significance.

关 键 词：片上可编程系统 图像去雾 暗通道先验算法 视频图像去雾 

分 类 号：TP368.2[自动化与计算机技术—计算机系统结构] TP751.1[自动化与计算机技术—计算机科学与技术]