文档介绍：运动目标识别技术在智能监控系统中的应用
----智能信息处理课程论文
摘要
本文首先综述了智能监控系统的发展历史和现状,然后着重分析了智能监控系统应用前景和发展趋势。接着阐述了视频图像预处理的必要性,图像预处理能有效地消除噪声,改善图像质量,使图像清晰化。中值滤波是一种非线性滤波方法,它不但对图像噪声消除非常有效,而且能够较好地保护图像边缘信息。本文用它来对视频图像进行预处理,取得了理想的处理效果,为目标的检测提供了良好的基础。在比较常用的几种运动目标识别方法之后,最终选用了背景差分法,因为这个方法可以直接得到运动对象的位置、大小、形状等信息,这些都是视频监控系统所需要的重要信息,并且背景差分法计算复杂度适中。因此本文采用背景差分法来实现运动目标的识别。实验仿真的结果表明,该方法简单实用,可以很容易地实现对运动目标的识别。
关键词:智能监控系统,图像预处理,运动目标识别,背景差分法
Abstract
At first this paper makes prehensive survey of history and the status of the art of intelligent monitoring system, then focuses on the application prospects and development trends of intelligent monitoring system. Then introduced the need for pre-processing video images, image pre-processing can effectively eliminate the noise and improve image quality and clarity of the image. Median filter is a nonlinear filtering method, it is not only very effective in the elimination of image noise, and can be used to protect the image edge information. In this paper, use it to carry out pre-processing of video images, and achieved the desired effect, provide a good foundation for target detection. After parison of monly used in moving target identification methods, this paper select the background difference method, because it can be directly targeted at the location, size, shape and other information, all these information are important in the video surveillance systems, and the background difference method plexity. Therefore, this paper uses background difference method to achieve the identification of moving targets. And the experimental simulation results show that this method is simple and practical, easy to implement.
Key words: Intelligent video surveillance systems, Image pre-processing, Moving target identification, The background difference method
1. 绪论
课题的研究背景及其意义
近年来,人们越来越关注生存环境,重视突发灾害的预带与紧急应对措施的建立,迫切希望提高生产和生活管理的智能化水平,视频监控系统也作为人类视觉的延伸越来越受到重视,得到了长足的发展[1]。
传统的视频监控系统由连接到一套电视监视器上的一个或多个摄像机组成。这种监控系统己经可以满足人们“眼见为实”的要求[2],但同时这种监控系统也存在一定的弊端[3]:
第一,它要求监控人员不停的监视