大数据技术在城域视频监控系统中的应用研究
发布时间：2019-01-15 08:58
【摘要】：随着社会和科技的快速发展,人们对于安全防范的意识也越来越高,政府部门对于人们的安全保障水平也需要快速提高,互联网技术,信息技术,视频压缩技术等的快速发展,使得城域视频监控系统也已经应用到了人们的日常工作和生活当中,视频监控对于保障人民生活和财产安全的作用是相当巨大的,但随着时间的推移,城域视频监控系统产生的数据量持续增长,必须利用新的数据存储技术、处理分析技术来适应数据量的快速增长。实际上不只是城域视频监控系统,各行各业都面临着大数据时代的到来,大数据给很多传统的处理分析技术带来了很大的挑战,为了能够适应大数据时代的需求,相应的数据处理分析技术开始快速发展,城域视频监控系统中同样需要引入这些大数据技术来分析处理持续增长的数据。本文通过对大数据相关技术的应用研究,利用云存储技术对城域视频监控系统产生的海量数据进行存储及管理,利用数据挖掘技术对存储的海量数据进行深入分析。首先利用MongoDB技术存储城域视频监控系统数据存储时的元数据信息,然后利用Hadoop分布式文件系统(HDFS)对原始数据进行存储,然后通过视频数据挖掘技术对原始监控视频进行挖掘,利用挖掘出的数据进一步深入分析,主要两个方面,一是嫌疑车辆侦查分析,利用视频监控中包含的信息设计了一套专家系统对案件嫌疑车辆进行识别,设计了系统知识库中的推理规则以及推理机制;二是通过采用关联规则的挖掘算法对道路交通事故成因进行分析,并采用信息论中的信息增益来选取事故的影响主因素。通过将大数据技术应用到城域视频监控系统中,可以更好的保障人民和社会的安全。
[Abstract]:With the rapid development of society and science and technology, people's awareness of security prevention is becoming higher and higher. The level of safety and security of government departments also needs to be improved rapidly. The rapid development of Internet technology, information technology, video compression technology, etc. The city video surveillance system has also been applied to people's daily work and life. Video surveillance plays a very important role in protecting people's life and property security, but with the passage of time, The amount of data generated by metro video surveillance system continues to grow, so it is necessary to use new data storage technology, processing and analysis technology to adapt to the rapid growth of data volume. In fact, it is not just the video surveillance system in the metropolitan area. All kinds of industries are facing the arrival of big data's era. Big data has brought great challenges to many traditional processing and analysis technologies, in order to be able to adapt to the needs of the era of big data. The corresponding data processing and analysis technology has been developing rapidly. It is also necessary to introduce these big data techniques to analyze and deal with the growing data in the metropolitan video surveillance system. In this paper, we use cloud storage technology to store and manage the mass data generated by the metropolitan video surveillance system through the application research of big data related technology, and use data mining technology to deeply analyze the mass data stored. At first, we use MongoDB technology to store metadata information in the data storage of metro video surveillance system, and then use Hadoop distributed file system (HDFS) to store the original data. Then through the video data mining technology to mine the original surveillance video, using the extracted data to further analyze the two main aspects, one is the suspect vehicle investigation analysis, A set of expert system is designed to identify the suspected vehicle by using the information contained in the video surveillance. The inference rules and reasoning mechanism in the system knowledge base are designed. Secondly, the cause of road traffic accidents is analyzed by mining association rules, and the main factors of accidents are selected by the information gain in information theory. By applying big data technology to the video surveillance system of metropolitan area, the security of people and society can be better guaranteed.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP311.13;TN948.6
本文编号：2409057
[Abstract]:With the rapid development of society and science and technology, people's awareness of security prevention is becoming higher and higher. The level of safety and security of government departments also needs to be improved rapidly. The rapid development of Internet technology, information technology, video compression technology, etc. The city video surveillance system has also been applied to people's daily work and life. Video surveillance plays a very important role in protecting people's life and property security, but with the passage of time, The amount of data generated by metro video surveillance system continues to grow, so it is necessary to use new data storage technology, processing and analysis technology to adapt to the rapid growth of data volume. In fact, it is not just the video surveillance system in the metropolitan area. All kinds of industries are facing the arrival of big data's era. Big data has brought great challenges to many traditional processing and analysis technologies, in order to be able to adapt to the needs of the era of big data. The corresponding data processing and analysis technology has been developing rapidly. It is also necessary to introduce these big data techniques to analyze and deal with the growing data in the metropolitan video surveillance system. In this paper, we use cloud storage technology to store and manage the mass data generated by the metropolitan video surveillance system through the application research of big data related technology, and use data mining technology to deeply analyze the mass data stored. At first, we use MongoDB technology to store metadata information in the data storage of metro video surveillance system, and then use Hadoop distributed file system (HDFS) to store the original data. Then through the video data mining technology to mine the original surveillance video, using the extracted data to further analyze the two main aspects, one is the suspect vehicle investigation analysis, A set of expert system is designed to identify the suspected vehicle by using the information contained in the video surveillance. The inference rules and reasoning mechanism in the system knowledge base are designed. Secondly, the cause of road traffic accidents is analyzed by mining association rules, and the main factors of accidents are selected by the information gain in information theory. By applying big data technology to the video surveillance system of metropolitan area, the security of people and society can be better guaranteed.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP311.13;TN948.6
【参考文献】
相关期刊论文 前10条
1 张锋军;;大数据技术研究综述[J];通信技术;2014年11期
2 涂新莉;刘波;林伟伟;;大数据研究综述[J];计算机应用研究;2014年06期
3 宋杰;郭朝鹏;王智;张一川;于戈;Jean-Marc PIERSON;;大数据分析的分布式MOLAP技术[J];软件学报;2014年04期
4 何清;李宁;罗文娟;史忠植;;大数据下的机器学习算法综述[J];模式识别与人工智能;2014年04期
5 张恩;张广弟;兰磊;;基于MongoDB的海量空间数据存储和并行[J];地理空间信息;2014年01期
6 张引;陈敏;廖小飞;;大数据应用的现状与展望[J];计算机研究与发展;2013年S2期
7 高汉松;肖凌;许德玮;桑梓勤;;基于云计算的医疗大数据挖掘平台[J];医学信息学杂志;2013年05期
8 宗群;李光宇;郭萌;;基于故障树的电梯故障诊断专家系统设计[J];控制工程;2013年02期
9 黄斌;许舒人;蒲卫;;基于MapReduce的数据挖掘平台设计与实现[J];计算机工程与设计;2013年02期
10 孟小峰;慈祥;;大数据管理:概念、技术与挑战[J];计算机研究与发展;2013年01期
相关硕士学位论文 前1条
1 黄晓云;基于HDFS的云存储服务系统研究[D];大连海事大学;2010年
，本文编号：2409057
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2409057.html
