电子学报 ›› 2013, Vol. 41 ›› Issue (11): 2307-2313.DOI: 10.3969/j.issn.0372-2112.2013.11.031

• 科研通信 • 上一篇    下一篇

基于ML和L2范数的视频目标跟踪算法

姜明新1,2, 王洪玉1, 王洁1, 王彪1   

  1. 1. 大连理工大学信息与通信工程学院, 辽宁大连 116024;
    2. 大连民族学院信息与通信工程学院, 辽宁大连 116600
  • 收稿日期:2012-11-28 修回日期:2013-04-09 出版日期:2013-11-25
    • 作者简介:
    • 姜明新 女,1979年出生于黑龙江省.大连理工大学信息与通信工程学院博士研究生,大连民族学院信息与通信工程学院讲师.研究方向为计算机视觉、智能视频监控 E-mail:jmx@mail.dlut.edu.cn 王洪玉 男,1968年出生.大连理工大学信息与通信工程学院教授,博士生导师.研究方向无线传感器网络、计算机视觉.
    • 基金资助:
    • 国家自然科学基金 (No.61172058); 中央高校自主基金 (No.DC10010103); 辽宁省教育厅资助项目 (No.L2012476)

Visual Object Tracking Algorithm Based on ML and L2-Norm

JIANG Ming-xin1,2, WANG Hong-yu1, WANG Jie1, WANG Biao1   

  1. 1. School of Information and Communication Engineering, Dalian University of Technology, Dalian, Liaoning 116024, China;
    2. School of Information and Communication Engineering, Dalian Nationalities University, Dalian, Liaoning 116600, China
  • Received:2012-11-28 Revised:2013-04-09 Online:2013-11-25 Published:2013-11-25
    • Supported by:
    • National Natural Science Foundation of China (No.61172058); Initiative Fund for the Central Universities (No.DC10010103); Supported by Education Department of Liaoning Province (No.L2012476)

摘要: 目标跟踪是计算机视觉领域的一个具有挑战性的问题,本文提出了一种基于ML(最大似然)估计和L2范数的视频目标跟踪算法.建立基于稀疏限制的ML模型,给样本中的异常像素分配较小的权值,减少异常像素对跟踪算法的影响.利用L2范数最小化进行稀疏编码求解.采用贝叶斯估计得出目标跟踪结果.与其他典型算法相比,本算法降低了计算的复杂度,对遮挡,旋转,尺度变化,光照变化等异常变化具有较强的鲁棒性.

关键词: 稀疏限制, 最大似然, L2范数最小化, 贝叶斯MAP估计

Abstract: The tracking of target is a challenging issue in computer vision.In this paper,we propose a visual object tracking algorithm based on ML estimation and L2-norm.Firstly,the model of sparsity constrained ML is established.Abnormal pixels in the samples will be assigned with low weights to reduce their affects on the tracking algorithm.Then,L2-norm minimization is used to solve the sparse coding.Finally,the object tracking results is obtained using Bayesian MAP estimation.Compared with other popular methods,our proposed method reduces the computational complexity and has stronger robustness to abnormal changes(e.g.occlusion,rotation,scale change,illumination,etc.)

Key words: sparsity constraint, maximum likelihood (ML), L2-norm minimization, Bayesian MAP(maximum a posteriori probability) estimation

中图分类号: