雾天道路监控系统视频图像增强算法的比较研究  

A Comparative Study on Fog Image Enhancement Algorithm for Intelligent Video Surveillance

作　　者：张道华 ZHANG Daohua(Bozhou College, Bozhou Anhui 236800, Chin)

机构地区：[1]亳州学院,安徽亳州236800

出　　处：《重庆科技学院学报（自然科学版）》2017年第3期102-105,共4页Journal of Chongqing University of Science and Technology：Natural Sciences Edition

基　　金：安徽省高等学校省级质量工程项目"大学生创客实验室建设计划‘星星之火’创客实验室"(2016ckjh124);安徽省高等学校省级质量工程项目"名师工作室"(2014msgzs170);安徽高校自然科学研究重点项目"基于敏捷开发的高校实习实训APP建设"(KJ2017A703)

摘　　要：随着平安城市的逐步推进,以智能视频监控为主流的图像增强技术出现了供应上的短缺。因此,研究降质图像的增强算法具有重要的科研价值和经济价值。对目前常用的雾天图像增强算法,如直方图均衡化算法、多尺度Retinex算法及暗通道优先算法,对其进行运行效率和算法效果的比较,并采用MATLAB分别实现3种算法处理后图片的查看和提取。实验表明,暗通道优先算法能够清晰地还原图像,但处理时间较长。最后对暗通道优先算法的进一步完善提出了研究方向。With the gradual advance of green ci ty, the supply of image enhancement technology for intel l igent video surveillance is inadequate to meet the demand. So the research of degraded image enhancement algorithm has im-portant research value and economic value. This article analyzes frequently - used fog image enhancement algorithm such as the histogram equalization algorithm, multi - scale Retinex and the dark channel priority algorithm, and compares the results and eff iciency of these three fog image processing algorithms. And then with the help of Matlab software, we view the final results of the three algorithmsr output. Experiments show that the dark channel priority algorithm can restore image clearly, but it costs much time. Lastly, we put forward further research direct ion about the dark channel priority algorithm.

关 键 词：图像增强 直方图均衡化 多尺度RETINEX 暗通道优先 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]