Multiple-Person Tracking by Multiple Cameras and Laser Range Scanners in Indoor Environments
Multiple-Person Tracking by Multiple Cameras and Laser Range Scanners in Indoor Environments
复制标题
DOI：
10.20965/jrm.2010.p0221
复制
发表时间：
2010-04-01
影响因子：
1.1
通讯作者：
Sato, Tomomasa
中科院分区：
文献类型：
--
作者：
Noguchi, Hiroshi;Mori, Taketoshi;Sato, Tomomasa
在本文中，我们提出了一种使用相机和激光距离扫描仪进行多人跟踪的方法。我们的方法估计人体和头部的三维位置，并标记他们的身份。单个粒子过滤器通过整合来自激光距离扫描仪的信息和来自摄像机的目标特定信息来正确地跟踪人，从而补偿各自的弱点。我们还开发了一个粒子滤波框架，利用估计的身体位置同时跟踪人类头部。实验结果证明了该方法在使用多台扫描仪和多台摄像机跟踪多人时的有效性和鲁棒性。
In this paper, we propose a method for multiple-person tracking using cameras and laser range scanners. Our method estimates 3D positions of human body and head, and labels them with their identities. Individual particle filters track person correctly by integrating information from laser range scanners and target-specific information from cameras, thus compensating for weak points of each. We also develop a particle filter framework that tracks the human head simultaneously using the estimated body position. Results of experiments demonstrate the effectiveness and robustness of the proposal in tracking multiple persons with multiple scanners and cameras.