计算机与现代化 ›› 2016, Vol. 0 ›› Issue (6): 12-15.doi: 10.3969/j.issn.1006-2475.2016.06.003

• 网络与通信 • 上一篇    下一篇

基于k均值的监控视频热度云存储研究

  

  1. 1.中国航天科工二院研究生院,北京100854;2.北京航天长峰科技工业集团有限公司,北京100854
  • 收稿日期:2015-12-09 出版日期:2016-06-16 发布日期:2016-06-17
  • 作者简介:张超(1990-),男,河南长垣人,中国航天科工二院研究生院硕士研究生,研究方向:数字图像处理; 郭会明(1965-),男,湖北天门人,北京航天长峰科技工业集团有限公司副总裁,硕士,研究方向:智慧城市。

Cloud Storage Research on Supervisory Video Heat Based on K-means

  1. 1. The Graduate School of Second Academy of China Aerospace, Beijing 100854, China;
    2.China Changfeng Science Technology Industry Group Corp, Beijing 100854, China
  • Received:2015-12-09 Online:2016-06-16 Published:2016-06-17

摘要: 通过对基于Hadoop的视频存储架构进行分析,提出一种面向智能化的双Hadoop集群存储架构。本文算法基于k均值预测监控视频热度,并根据平安城市监控特点对视频备份数进行调整,从而可提高设备利用率,加快视频文件检索以及智能化分析速度。

关键词: 视频监控, 热度, 智能化, 云存储

Abstract: Through the analysis of Hadoop video storage structure, the article comes up with an intelligent double Hadoop cluster storage architecture. The algorithm predicts the supervisory video heat based on k-means, and according to the features of safe city adjusts the number of backups. As the result, this method can increase the use ratio of facilities, accelerate the speed of document retrieval and intelligence analysis.

Key words:  video surveillance, heat, intelligence, cloud storage

中图分类号: