{
  "paper_id": "b97d677b9dc8d879e9bbb0cff9ff4c11",
  "paper_title": "People Tracking in RGB-D Data With On-line Boosted Target Models",
  "paper_authors": [
    "Matthias Luber",
    "Luciano Spinello",
    "Kai O. Arras"
  ],
  "publication_date": "",
  "publication_venue": "2011 IEEE/RSJ International Conference on Intelligent Robots and Systems",
  "abstract": "People tracking is a key component for robots that are deployed in populated environments. Previous works have used cameras and 2D and 3D range finders for this task. In this paper, we present a 3D people detection and tracking approach using RGB-D data. We combine a novel multi-cue person detector for RGB-D data with an on-line detector that learns individual target models. The two detectors are integrated into a decisional framework with a multi-hypothesis tracker that controls on-line learning through a track interpretation feedback. For on-line learning, we take a boosting approach using three types of RGB-D features and a confidence maximization search in 3D space. The approach is general in that it neither relies on background learning nor a ground plane assumption. For the evaluation, we collect data in a populated indoor environment using a setup of three Microsoft Kinect sensors with a joint field of view. The results demonstrate reliable 3D tracking of people in RGB-D data and show how the framework is able to avoid drift of the on-line detector and increase the overall tracking performance.",
  "keywords": [
    "people tracking",
    "RGB-D data",
    "on-line boosting",
    "multi-hypothesis tracker",
    "person detection",
    "3D tracking"
  ],
  "paper_institutions": [],
  "output_language": "zh",
  "provider": "doubao",
  "model": "ark-code-latest",
  "prompt_template": "simple",
  "extracted_at": "2026-08-16T19:46:58.983120Z",
  "template_tag": "simple",
  "summary": "### 摘要\n本文针对机器人部署在人口密集环境中的关键问题——人群跟踪，提出了一种基于RGB-D数据的3D人物检测与跟踪方法。该方法结合了新型多线索人物检测器（Combo-HOD）与在线学习个体目标模型的检测器，将两者集成到一个决策框架中，由多假设跟踪器（MHT）通过轨迹解释反馈控制在线学习。在线学习采用AdaBoost方法，使用三种类型的RGB-D特征（Haar-like特征在强度图、深度图以及Lab颜色特征）和在3D空间中的置信度最大化搜索。方法具有通用性，不依赖背景学习或地面平面假设。在评估中，使用三个Microsoft Kinect传感器联合视场，在人口密集的室内环境（大学食堂大厅）收集数据。结果表明，该方法能够在RGB-D数据中可靠地跟踪3D人物，并且框架能避免在线检测器的漂移，提高整体跟踪性能。核心机制为在线增强用于特征选择和置信度最大化搜索，结合联合似然数据关联（运动与外观）来引导关联，并通过轨迹解释反馈控制学习与暂停，以避免模型漂移。\n\n### 关键要点\n- 提出了一种新颖的RGB-D数据3D人物检测与跟踪框架，结合先验检测器（Combo-HOD）和在线学习的目标模型检测器。\n- 在线检测器采用在线AdaBoost方法，使用Haar-like特征（强度图、深度图）和Lab颜色特征，通过特征选择器动态选择最佳弱分类器。\n- 置信度最大化搜索在3D空间中进行，搜索窗口大小与运动预测不确定性成正比，优于固定大小搜索窗口。\n- 多假设跟踪器（MHT）基于Kalman滤波，处理轨迹生命周期，包括创建、确认、遮挡和删除。\n- 联合似然数据关联结合了运动似然（高斯模型）和外观似然（基于强分类器置信度的高斯模型），以改善数据关联。\n- 轨迹解释反馈控制在线学习：新目标时初始化分类器，关联时更新，遮挡时暂停学习并搜索替代观测，避免漂移。\n- 实验设置：三个垂直安装的Kinect传感器，联合视场为130°×50°，分辨率为1440×640像素，以30Hz同步获取数据。\n- 数据集在午餐时间大学食堂大厅采集，包含1133帧、3021个目标实例、31条轨迹，手动标注了2D边界框和可见性及数据关联真值。\n- 使用CLEAR MOT指标评估，基线仅用先验检测器，对比在线增强方法。结果显示漏检数减少50%，标识切换减少24%，MOTA从62%提升至78%（提升16%）。\n- 在线增强方法导致误报增加19%，主要由于对静态物体上重复的错误检测在线学习适配良好，导致轨迹延迟删除。\n- 在线检测器在遮挡期间暂停学习，重新出现后恢复高置信度，有效避免漂移。\n- 未来工作将收集更多具有挑战性社交场景的RGB-D数据集，进一步验证方法的鲁棒性。"
}