作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2007, Vol. 33 ›› Issue (21): 24-27. doi: 10.3969/j.issn.1000-3428.2007.21.009

• 博士论文 • 上一篇    下一篇

基于均值漂移的视觉目标跟踪方法综述

齐 飞,罗予频,胡东成   

  1. (清华大学自动化系,北京 100084)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-11-05 发布日期:2007-11-05

Overview on Visual Target Tracking Based on Mean Shift

QI Fei, LUO Yu-pin, HU Dong-cheng   

  1. (Department of Automation, Tsinghua University, Beijing 100084)
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-11-05 Published:2007-11-05

摘要: 基于均值漂移的视觉目标跟踪方法具有模型简洁实用、能够处理目标形变及部分遮挡等复杂情形的优点,算法高效且易于模块化实现。各种改进的模型及方法针对目标的尺度变化、特征分布等核心问题进行了系统研究,跟踪性能得到了进一步提高。该文从基本的均值漂移跟踪方法出发,系统介绍了此类方法的发展过程与最新成果。

关键词: 均值漂移, 视觉目标跟踪, 核函数, 相似性度量

Abstract: Mean-shift-based visual target tracking is one of the hotspots in the field of computer vision. The model of the algorithm is simple, efficient and easy-to-implement, and it can handle the complex cases such as deformations and partial occlusions. Recent researches on scale adaptation of the tracking window and distributions of features improve the performance of such trackers. This paper introduces the development and the state of such kind of the algorithms.

Key words: mean shift, visual target tracking, kernel functions, similarity measurement

中图分类号: