{
  "paper_id": "c47ab29d722a09afb1b67b29f4ea9ebd",
  "paper_title": "Robust Human Tracking Using a 3D LiDAR and Point Cloud Projection for Human-Following Robots",
  "paper_authors": [
    "Sora Kitamoto",
    "Yutaka Hiroi",
    "Kenzaburo Miyawaki",
    "Akinori Ito"
  ],
  "publication_date": "2025-03-12",
  "publication_venue": "Sensors",
  "abstract": "Human tracking is a fundamental technology for mobile robots that work with humans. Various devices are used to observe humans, such as cameras, RGB-D sensors, millimeter-wave radars, and laser range finders (LRF). Typical LRF measurements observe only the surroundings on a particular horizontal plane. Human recognition using an LRF has a low computational load and is suitable for mobile robots. However, it is vulnerable to variations in human height, potentially leading to detection failures for individuals taller or shorter than the standard height. This work aims to develop a method that is robust to height differences among humans using a 3D LiDAR. We observed the environment using a 3D LiDAR and projected the point cloud onto a single horizontal plane to apply a human-tracking method for 2D LRFs. We investigated the optimal height range of the point clouds for projection and found that using 30% of the point clouds from the top of the measured person provided the most stable tracking. The results of the path-following experiments revealed that the proposed method reduced the proportion of outlier points compared to projecting all the points (from 3.63% to 1.75%). As a result, the proposed method was effective in achieving robust human following.",
  "keywords": [
    "3D LiDAR",
    "human tracking",
    "human-following robot"
  ],
  "paper_institutions": [
    "Graduate School of Robotics and Design, Osaka Institute of Technology",
    "Faculty of Robotics and Design, Osaka Institute of Technology",
    "Faculty of Information Sciences and Technology, Osaka Institute of Technology",
    "Graduate School of Engineering, Tohoku University"
  ],
  "output_language": "zh",
  "provider": "doubao",
  "model": "ark-code-latest",
  "prompt_template": "simple",
  "extracted_at": "2026-08-16T19:46:50.325008Z",
  "template_tag": "simple",
  "summary": "### 摘要\n本文提出一种基于3D LiDAR的人体跟踪方法，旨在解决传统2D LRF因人体高度差异导致的检测不稳定问题。相关工作表明，2D LRF仅观测单一水平面，对身高变化敏感；而3D LiDAR能获取多平面点云，但直接处理点云计算成本高。本方法将3D点云投影到2D平面，从而复用已有的2D LRF人体跟踪算法，具体通过公式 r̂(θ)=min_{φ∈V} r(θ,φ)cosφ 选择投影范围，其中V由从头顶向下取固定比例（robs）的点云决定。实验使用Velodyne VLP-16-LITE，在机器人上安装于1.2m高度，11名参与者（身高1.58-1.73m）进行身高测量，平均RMS误差3.83cm；通过路径跟随实验比较不同robs（0.2,0.3,0.4,0.5），发现robs=0.3时人体宽度估计最稳定，方差最小。进一步在10条路径上对比robs=0.3与全点云（robs=1）的效果，以连续观测点距离超过0.2m为离群点，结果表明robs=0.3的离群点比例从3.63%降至1.65%，具有统计显著性（p=0.0055）。该方法计算高效，点云投影、人体检测和跟踪分别耗时1.49ms、0.574ms和0.111ms，总处理2.175ms，无需GPU。未来工作将扩展至多人场景、复杂环境，并结合传感器融合提升鲁棒性。\n\n### 关键要点\n- 传统2D LRF因观测平面固定，对身高较高或较矮的人体会产生检测错误，如捕获手臂或头部。\n- 本方法将3D LiDAR点云投影到水平面，采用最近点策略生成2D扫描，以复用已有的2D LRF跟踪算法。\n- 投影范围由robs决定，即从头顶向下取人体高度的比例，实验优化得到robs=0.3为最优。\n- 身高测量中，参与者站在2.5m处，通过最高检测仰角确定身高，平均RMS误差为3.83cm。\n- 人体宽度稳定性实验表明，robs=0.2时宽度明显偏窄（测量头部），而robs≥0.3时测量的是身体宽度。\n- 方差分析显示不同robs下的宽度方差有显著差异（Barlett检验p=1.93×10^-10），F检验将0.2、0.3与0.4、0.5分为两组。\n- 在10条路径（L型、S型、Z型等）的实验中，robs=0.3的离群点比例（1.65%）显著低于全点云（3.63%）单侧t检验p=0.0055。\n- 手臂运动是离群点主因，全点云投影时手臂被误判为身体，导致宽度和位置波动。\n- 计算效率高：点云投影1.49ms，人体检测0.574ms，跟踪0.111ms，总计2.175ms，无需GPU。\n- 方法假设单一预注册目标，且注册时人需站在固定距离（2.5m），这限制了实际应用。\n- 未来计划处理多人场景、更复杂环境，并整合运动模型和传感器融合以进一步提升鲁棒性。"
}