摘要近几十年来，在科学技术的飞速发展下，视频与图像处理技术不断发展提高，各式各样的视频监控产品已经走入了人们的视野，并且给我们生活的方方面面带来很多便利。智能视觉监控系统就是一项利用计算机对视频图像进行处理和分析，能够代替人眼，具有自动捕捉、识别目标功能的技术产品，并且已经在现实生活中得到了越来越广泛的应用。例如交通方面可用于车辆的监控与追踪，公众场合可用于防盗监控等。而其所面临的最基本的问题就是视频序列图像中对运动目标的检测。78807

本文分别采用了帧差法、光流法和背景差分法进行视频序列的运动目标检测。其中，帧差法分为了一般帧差法和对称帧差法，分别取相邻两帧和三帧图像进行差分研究，从中对比各自优缺点。光流法整体较复杂，在本文中做了适当简化，仅选取了一种颜色系统来显示。背景差分法则分别采用了平均法建模和单高斯模型建模，分别得到相应的结果并对比分析。通过几种方法处理视频后的结果对比，从而找出不同场合下最优的运动目标检测方法。

毕业论文关键词  视频运动目标检测      帧差法      光流法     背景差分法

毕业设计说明书外文摘要

Title        Method of moving object detection in vedio             

Abstract  Various video surveillance products have gone into people’s vision and bring in much convenience in almost every aspect of our life for decades,with the rapid development of science and technology and the continuous development of processing technology for videos and images。 Intelligent visual surveillance system is a technological product with the function of automatically capturing and identifying targets which mainly uses computers to process and analyze video images and has been used more and more widely in real life。For example,it can be used to monitor and track vehicle in transportation and monitor to guard against theft in public。The detection of moving objects in video sequence images is the most fundamental problem that confronts the system。

    The frame difference method,the optic flow method and the background subtraction method are adopted in this paper to detect the moving objects。Above all,the frame difference method includes general frame difference method and symmetrical frame difference method。And the two methods study the adjacent frames and the three frame difference images respectively。This paper makes a comparison between the two methods in this process。The optional flow method is complicated on the whole,so this paper makes some appropriate predigestion and choose only one color system to display this method。 The averaging model and the single gauss model are used respectively in the background subtraction method。This paper make a comparative analysis of the corresponding result between the two models。This paper make a contrast between the results which obtain from processing videos by several ways to find out the optimal way for detecting moving objects in different situations。

Keywords  Video moving object detection   optical flow method   background- difference method     frame difference method

目   次

1  绪论 1

1。1  研究背景 1

1。2  研究的现实意义 1

1。3  研究现状和所存在的问题 2

1。4  各章节安排 3

2  研究方法介绍