基于时空兴趣点的化工厂视频监控系统设计与实现

发布时间：2018-03-20 19:23

本文选题：视频监控系统　切入点：时空兴趣点　出处：《电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：最近这几年,监控摄像头在生活中各个领域迅速普及,视频数据量成倍增加,传统的依靠人工的监控方式其效率和准确性难以保证,这使得智能视频监控成为当今计算机视觉领域的一个研究热点。而运动目标的检测与跟踪是视频分析的关键内容,在视频监控、军事应用、人机交互等很多方面都是应用实现的基础,是计算机视觉中的活跃领域。在化工厂这样的环境下,一旦发生事故,比如起火或者认为破坏等,将造成严重的后果,因此监控设备就尤为重要,在一些敏感位置都安装了摄像头,如果能够实时地对这些监控场景进行分析,跟踪和判断场景中人的行为,对一些异常的现象发出警报,就能防患于未然,确保安全。出于以上目的,本文自主研制和设计了一套稳定的、高效的、人性化的视频监控系统。本系统的智能检测模块主要应用了基于时空兴趣点的检测方法。本文中的新的检测算法,在准确度上有了较大的提高,使智能检测模块的判断更加准确。当监控现场出现异常情况时,检测模块能通过视频迅速检测到,并以最快和最佳的方式发出报警,尽最大可能的杜绝人为因素造成的事故,在功能上能够实现事故发生前发出报警,保证人员在事故中的正确处理以及事故后能完全取证的智能监控系统。考虑到本文需要存储的数据量相当大,所以在数据库建立方面进行了大量的研究和尝试。以ER模型为基础,总结出了适合本系统的数据存储方式。即建立多个数据库用来存储不同的数据,同时详细设计了一部分数据表,同时还从安全方面考虑,加强了对数据库的安全管理。本文还主要介绍了监控系统的软件设计与开发,包括客户端软件设计和服务器端软件设计,其中重点是对客户端软件的设计。通过深入研究系统的需求关系,本文将软件的模块分为:登陆模块,用户管理模块,报警查询模块,录像回放模块等多个功能模块。
[Abstract]:In recent years, surveillance cameras have been rapidly popularized in all fields of life, and the amount of video data has multiplied. It is difficult to guarantee the efficiency and accuracy of traditional surveillance methods relying on manual methods. This makes intelligent video surveillance a research hotspot in the field of computer vision, and the detection and tracking of moving targets is the key content of video analysis, in video surveillance, military applications, Many aspects, such as human-computer interaction, are the basis of application and implementation, and are active fields in computer vision. In a chemical plant such as an environment, once an accident occurs, such as fire or damage, it will have serious consequences. So monitoring equipment is particularly important. Cameras are installed in some sensitive locations. If you can analyze, track and judge the behavior of people in the scene in real time, you can warn of some unusual phenomena. For the above purpose, this paper independently developed and designed a set of stable, efficient, Humanized video surveillance system. The intelligent detection module of this system mainly uses the detection method based on space-time interest point. The new detection algorithm in this paper has greatly improved the accuracy. Make the judgment of intelligent detection module more accurate. When there is abnormal situation in the monitoring scene, the detection module can quickly detect through the video, and send out the alarm in the fastest and best way, so as to prevent the accidents caused by human factors as far as possible. The function can realize the alarm before the accident occurs, ensure the correct handling of the personnel in the accident and the intelligent monitoring system which can collect the evidence completely after the accident. Considering the large amount of data needed to be stored in this paper, So in the aspect of database establishment, a lot of researches and attempts have been made. On the basis of ER model, the data storage method suitable for this system has been summarized. That is to say, many databases have been established to store different data. At the same time, some data tables are designed in detail, at the same time, the security management of database is strengthened from the aspect of security. This paper also mainly introduces the software design and development of monitoring system. Including client software design and server-side software design, the emphasis is on the client software design. Through the in-depth study of the requirements of the system, this paper divides the software modules into: landing module, user management module, Alarm query module, video playback module and other functional modules.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN948.6

【参考文献】