论文总字数：17590字

摘 要

随着现代社会以及高科技的飞速发展，在社会安保系统中，视频监控的作用越来越受到重视。而在现实生活中由专人看着屏幕来实现监控的系统已经远不能达到现代安保的要求，于是智能视觉监控系统就出现了。智能视觉监控技术，指的是只通过将智能算法嵌入DSP中来用计算机等设备来对摄像机所拍的视频中的物体进行运动目标的检测、跟踪甚至定位等的各种行为分析的技术，而不是通过有人参与其中进行分析的技术。而其中对运动目标的检测、跟踪方面的技术则是智能视觉监控系统中最根本的也是最基础的技术。

本文先简单介绍了目标定位与追踪研究的背景与意义以及在国内外的研究现状与发展前景。因为本文是基于OpenCV来完成的课题，所以之后对OpenCV进行了简要说明已经主要功能的介绍。然后介绍了在运动目标的检测方法中最常见的三个算法，有帧间差分法、光流法以及背景差分法。

然后在实现运动目标的定位与追踪的实验里，本文选用了背景差分法与Camshift算法。详细介绍了使用背景差分法进行运动目标检测的流程然后进行分析，再对Camshift算法如何实现对运动目标的追踪进行详细介绍。

最后，在使用Visual Studio软件配置好opencv环境的前提下，对本文所说的算法进行研究实验的验证，在失败了几次之后最终取得了成功的实验结果。

关键词：目标检测 目标跟踪 CpenCV 背景差分法 Camshift算法

Abstract

With the rapid development of modern society and high technology, the role of video surveillance has received increasing attention in social security systems. In real life, the system that is monitored by a person watching the screen is far from meeting the requirements of modern security. Therefore, the intelligent visual monitoring system has emerged. The intelligent visual monitoring technology refers to the technology that analyzes, analyzes, and even locates the moving objects such as computers, etc., by using a computer or other device to capture, track, or even locate the objects in the video captured by the camera. It is not through the technology in which people are involved in the analysis. Among them, the technology of detection and tracking of moving targets is the most fundamental and basic technology in intelligent visual monitoring systems.

This paper first briefly introduces the background and significance of the research on target positioning and tracking, and the research status and development prospects at home and abroad. Because this article is based on OpenCV to complete the task, so after the OpenCV brief description of the main functions have been introduced. Then the three most common algorithms in moving object detection methods are introduced, including the inter-frame difference method, the optical flow method, and the background difference method.

Then in the experiment of realizing the positioning and tracking of moving targets, this paper chooses the background difference method and Camshift algorithm. The process of using the background difference method to detect the moving target is introduced in detail and then analyzed, and then the Camshift algorithm is introduced in detail about how to track the moving target.

Finally, under the premise of using the Visual Studio software to configure the opencv environment, the verification experiment of the algorithm described in this paper was performed. After a few failures, the successful experimental results were finally obtained.

Keywords： target detection; target tracking，CpenCV，background difference method，Camshift algorithm

目录

摘要 I

Abstract II

第一章 绪论 1

1.1目标定位与追踪研究的背景和意义 1

1.2 目标跟踪技术在国内外的研究及应用概述 2

1.3 目标定位与追踪技术的难点 3

第二章 OpenCV 4

2.1 OpenCV简介 4

2.2 opencv的功能 4

2.3 opencv模块 5

第三章 运动目标在静态背景下的检测方法 6

3.1 帧间差分法 6

3.2背景差分法 6

3.3光流场法 7

3.4本章小结 8

第四章 基于背景差分法的目标检测方法 9

4.1 背景建模 9

4.2 二值化 9

4.3 形态学滤波处理 10

4.4 实验结果展示 11

4.5 本章小结 13

第五章 基于Camshift算法的运动目标跟踪 14

5.1 Camshift算法介绍 14

5.2 Camshift 的OpenCV实现 19

5.3 实验结果与分析 22

5.4本章小结 23

第六章 总结与展望 24

参考文献 25

致谢 26

请支付后下载全文，论文总字数：17590字