摘要:随着国民经济的飞速发展，人们对楼宇智能化的要求越来越高，楼宇监控系统的应用也越来越广泛，需要监控的对象越趋多样化。由于传统的视频监控系统存在功能单一，存储空间占用大，运行效率低的缺点，为适应市场需要，提高产品竞争力，需要对原始视频进行智能提取，利用计算机的强大运算能力实现智能化的视频监控。

　　　本文先介绍了图像预处理的相关内容及运动目标检测关键技术，分析了运动目标检测算法。在这些基础上，主要通过实验针对智能监控中的运动目标，运用MATLAB软件完成实验，提出用于背景差的视频序列的运动目标快速检测算法，实验宇通道视频记录的各帧图像进行目标检测，将5小时的冗长的楼宇通道计算机监控视频文件浓缩为2小时精炼、准确、连续的进出状况专门视频文件。 

关键词：智能提取；视频分析；运动目标检测；MATLAB

Abstract:With the rapid development of national economy, people now have higher requirements for the quality of life. The pace of life is faster and faster and the pressure of work today is larger and larger. people have a strong desire to live a stable and comfortable live, which greatly promoted the development of intelligent community. As there are some disadvantage on the traditional video surveillance system, such as single function, large storage space occupation, running low efficiency. To meet the market requirements, enhance product competitiveness, it's necessary to intelligently extract the original video, to achieve intelligent video surveillance by the powerful computer processing ability.

    First of all ,this paper introduces the related content of image preprocessing and moving object detecting key technology, analyzes the target motion detection algorithm is proposed.On these bases, this paper, in view of the moving object intelligent monitoring, Using MATLAB software to finish the experiment.Puts out the fast detection algorithm of moving object for video sequence of still background, conduct the target detection for each frame of the lengthy building channel computer monitor video files of 5 hours, then make it into a refined , accurate and continuous situation specifically video file of 2 hours .

Key words：Intelligent Extraction ; video analysis ; moving target detection ; MATLAB

　　　基于背景差分析方法的视频监控系统是结合计算机技术、数字图像处理技术、多媒体技术和工业电视技术的综合应用模式，具有性能优越，操作方便、对监控状态判断准确等的特点。在日常生活许多环境下能发挥其作用。并且越来越被社会所接受。如:智能楼宇视频分析监控、智能交通系统、重要地区进出口监控系统等。它的使用将会减少日常生活中的监控压力，为出现意外提供有效的证据，促进了生产和工作的效率。给社会带来了直接的经济效益，是未来的一种理想的监控模式。