摘要
齐齐哈尔市公安视频监控系统中每天所产生的大量视频数据对视频图像的检索、管理及安全产生了迫切的需求,视频图像的检索存在两个急需解决的问题,一是视频检索的准确度问题,二是检索效率的问题。面对海量的视频数据库,提出了基于Map/Reduce分布式计算模型与关键帧算法结合的方法,既提高了检索效率,又提高了检索准确率。
With the rapid development of Internet and multimedia technology, city public security video monitoring system produce a large number of video data every day, many problems need to be solved as video image retrieval, management and safety of the pressing demand. In general there are two points, one is the video retrieval accuracy problem, the other is the issue of retrieval efficiency. In the face of huge amounts of video database,the method is proposed based on the Map/Reduce distributed computing model combined with algorithm of key frames, which improve the efficiency and accuracy of retrieval.
出处
《电视技术》
北大核心
2015年第4期33-36,共4页
Video Engineering
基金
科技部科技惠民计划项目(2013GS230301)
