期刊文献+

单目视觉多目标跟踪方法的数据关联策略综述 被引量:1

Survey of Data Association Strategies in Monocular Vision Multi-Target Tracking
下载PDF
导出
摘要 多目标跟踪作为一项被广泛应用于视频监控、交通路况控制以及自动驾驶等领域的计算机视觉技术,深受相关研究人员的青睐。基于检测的多目标跟踪方法成为如今主流的做法,即对跟踪目标进行识别检测,然后运用数据关联策略生成目标轨迹。在现实生活环境中,场景是非常复杂且跟踪目标的数量和类别通常是不确定的,因此数据关联策略在多目标跟踪方法中是非常重要的环节。本文重点对单目视觉多目标跟踪过程中的数据关联策略进行了综述,系统地介绍了多目标跟踪中的数据关联策略。首先,描述数据关联的研究现状以及对多目标跟踪进行了概述;其次,对数据关联的概念及其需要解决的问题进行了详细介绍;然后,对各种数据关联策略进行了分析总结;最后,对多目标跟踪的数据关联策略的研究方向进行了展望。 As a computer vision technology widely used in video surveillance,traffic condition control and automatic driving,multi-tar⁃get tracking is favored by related researchers.The multi-target tracking method based on detection has become the mainstream prac⁃tice nowadays,that is,the tracking target is identified and detected and then the target trajectory is generated by using the data associa⁃tion strategy.In the real life environment,the scene is often very complex and the number and category of tracking targets are often un⁃certain,so the data association strategy is a very important link in the multi-target tracking method.This paper focuses on the overview of the data association strategy in the process of monocular vision multi-target tracking and systematically introduces the data associa⁃tion strategy in the process of multi-target tracking.Firstly,the research status of data association and multi-target tracking are de⁃scribed.Secondly,the concept of data association and the problems to be solved are introduced in detail.Then,various data association strategies are analyzed and summarized.Finally,the research direction of multi-target tracking data association strategy is prospected.
作者 柴文光 陈香远 Chai Wenguang;Chen Xiangyuan(School of Computers,Guangdong University of Technology,Guangzhou 510006)
出处 《现代计算机》 2021年第23期112-118,123,共8页 Modern Computer
关键词 多目标跟踪 数据关联 JPDA 网络流 匈牙利算法 端对端方法 multi-target tracking data association JPDA net slow hungarian algorithm the end-to-end method
  • 相关文献

参考文献8

二级参考文献21

共引文献59

同被引文献2

引证文献1

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部