›› 2015, Vol. 28 ›› Issue (12): 45-.

• 论文 • 上一篇    下一篇

基于粒子滤波的TLD目标跟踪算法

谷文华,蒋念平,李振兴   

  1. (上海理工大学  光电信息与计算机工程学院,上海 200093)
  • 出版日期:2015-12-15 发布日期:2015-12-15
  • 作者简介:谷文华(1989—),女,硕士研究生。研究方向:图像处理,自动化控制。E-mail:gwh591854006@163.com。蒋念平(1957—),男,副教授。研究方向:计算机应用,仪器仪表开发等。

TLD Target Tracking Algorithm Based on Particle Filter

GU Wenhua,JIANG Nianping,LI Zhenxing   

  1. (School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)
  • Online:2015-12-15 Published:2015-12-15

摘要:

在复杂背景下运动目标速度快速变化、目标丢失、或大面积遮挡时,TLD算法跟踪的稳定性明显下降,跟踪效果差。针对这一问题,在TLD算法的跟踪模块中引入粒子滤波,使其在长期稳定跟踪的基础上,更好地完成复杂背景下快速运动目标的跟踪。实验以OpenCV和VS2010为测试环境,结果表明,改进后的TLD算法较原始TLD算法能够更好地跟踪快速运动的目标,增强了跟踪的稳定性和鲁棒性。

关键词: 目标跟踪, TLD算法, 粒子滤波, OpenCV VS2010

Abstract:

The stability and tracking performance of TLD tracking algorithm decreased significantly with rapid changes in speed moving targets,goals lost in a complex background,or large shelter.The particle filter is introduced to the TLD algorithm tracking module to offer better complex background track fast-moving target as well as long-term stability.Experiments in OpenCV and VS2010 show that the improved TLD algorithm can better track fast-moving target,increasing the tracking stability and robustness.

Key words: target tracking;particle filter;TLD algorithm;OpenCV VS2010

中图分类号: 

  • TP391.41