中国科学院光电技术研究所机构知识库
Advanced  
IOE OpenIR  > 光电技术研究所博硕士论文  > 学位论文
题名:
基于均值平移算法的运动目标跟踪技术研究
作者: 刘素珍
学位类别: 硕士
答辩日期: 2008-06-04
授予单位: 中国科学院光电技术研究所
授予地点: 光电技术研究所
导师: 邓和林
关键词: 目标跟踪 ; 均值平移 ; 特征融合 ; kalman预测 ; 实时系统
其他题名: The Research on Moving Object Trackinig based on Mean-Shift Algorithm
学位专业: 信号与信息处理
中文摘要: 运动目标跟踪是计算机视觉处理中的一个热点,有着非常广泛的应用前景;同时视觉环境的多样性和复杂性使其成为图像处理领域的难点。本文以均值平移算法为核心算法,并针对其固有缺陷作了较好的改进,实现了复杂背景下的目标跟踪。 均值平移算法以核函数加权下的灰度直方图描述目标特征,当目标灰度和背景灰度相近时,可能导致跟踪失败,为此提出了一种基于直方图特征融合的均值平移目标跟踪方法;针对传统均值平移算法不能自适应于目标尺度的变化,结合尺度空间信息量度量方法和均值平移算法,较好的解决了目标尺寸变大或变小时的目标跟踪问题。 利用Kalman滤波器的预测功能,对目标的状态进行有效预测,以Kalman预测结果作为均值平移搜索的起始点,经过若干次匹配便可得到目标的准确位置,提高了算法的运算速度。另外,通过kalman滤波残差的大小对目标是否发生遮挡进行判断,当目标发生遮挡时,kalman滤波器停止工作,根据运动系统参数的估计值通过状态方程对目标后续状态进行估计,等待目标的再次出现,避免了目标被大比例遮挡带来的跟踪任务的失败。 本文搭建了实时运动目标跟踪实验平台,实现了视频图像采集和视频显示等功能。并且在实验平台上实现了基于传统均值平移算法的目标跟踪,在多次实验调试下,得到了较好的跟踪效果。
英文摘要: Moving object tracking has received considerable attention in computer vision due to its wide application prospect. Meanwhile the diversity and complexity of visual environment make it difficult to track the object effectively. This paper complete the tracking task under complex background based on Mean-Shift algorithm and some improved algorithm which is aimed at the inherent defects of Mean-Shift. Mean-Shift algorithm describes the feature of target using the distribution of gradient histogram which is weighted by a kernel function. When the gradient of target is closed to the gradient of background, it is possible to track the object unsuccessfully. To overcome this disadvantage, a new Mean-Shift tracking scheme that is based on histogram fusion is proposed. As traditional Mean-Shift algorithm cannot achieve object tracking effectively when the size of target change, we propose that Mean-Shift algorithm, combined with information measure of scale-space, can track the object excellently in the scenarios of not only increasing object scale but also decreasing object scale. It is feasible to effectively forecast the state of moving object using the pre-estimating function of Kalman Filter. The forecasting result of Kalman Filter can be the starting point of Mean-Shift, then we can get the precise target location after several matching operation, and the speed of algorithm can be improved. In addition, whether the object is occluded is judged by the difference of Kalman Filter. When the object is occluded, Kalman Filter is stopped, and the system estimate the subsequent state of object based on the estimating value of system parameter and state function, thus avoid the failure of target tracking which is brought by object occlusion. Real-time object tracking platform is constructed, and such basic functions as video image capture and video display are realized on it. The tracking algorithm based on traditional Mean-Shift is carried out on the platform, and we get preferable tracking effect during numerous experimental tests.
语种: 中文
内容类型: 学位论文
URI标识: http://ir.ioe.ac.cn/handle/181551/325
Appears in Collections:光电技术研究所博硕士论文_学位论文

Files in This Item:
File Name/ File Size Content Type Version Access License
10001_200528015127041刘素珍_paper.doc(3855KB)----限制开放View 联系获取全文

Recommended Citation:
刘素珍. 基于均值平移算法的运动目标跟踪技术研究[D]. 光电技术研究所. 中国科学院光电技术研究所. 2008.
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[刘素珍]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[刘素珍]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
文件名: 10001_200528015127041刘素珍_paper.doc
格式: Microsoft Word
此文件暂不支持浏览
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2016  中国科学院光电技术研究所 - Feedback
Powered by CSpace