{
  "paper_id": "26720452670b8fcbc46be9f97239e3f4",
  "paper_title": "3D-Sensing Distributed Embedded System for People Tracking and Counting",
  "paper_authors": [
    "Andres Burbano",
    "Samir Bouaziz",
    "Marius Vasiliu"
  ],
  "publication_date": "",
  "publication_venue": "2015 International Conference on Computational Science and Computational Intelligence (CSCI)",
  "abstract": "The present study focuses on the development of an embedded smart camera network dedicated to track and count people in public spaces. In the network, each node is capable of sensing, tracking and counting people while communicating with the adjacent nodes of the network. Each node typically uses a 3D-sensing camera positioned in a downward-view but the designed framework can accept other configurations. We present an estimation method for the relative position and orientation of the depth cameras. This system performs background modeling during the calibration process, using a fast and lightweight segmentation algorithm.",
  "keywords": [
    "people tracking",
    "embedded system",
    "3D-sensing",
    "background modeling",
    "scene analysis"
  ],
  "paper_institutions": [
    "IEF, Université Paris-Saclay",
    "R&D, Shopline"
  ],
  "output_language": "zh",
  "provider": "doubao",
  "model": "ark-code-latest",
  "prompt_template": "simple",
  "extracted_at": "2026-08-16T19:46:11.577267Z",
  "template_tag": "simple",
  "summary": "### 摘要\n本论文聚焦于开发用于公共空间人员跟踪与计数的嵌入式智能相机网络。每个网络节点配备3D深度相机、嵌入式计算机和通信单元，可本地检测、跟踪和计数人员，并与相邻节点通信。针对人群遮挡、人员粘连、阳光干扰等问题，系统采用俯视安装的深度相机，利用深度传感器对光照不敏感的优势，并设计了基于背景建模的快速分割方法。系统使用Asus Xtion Pro或Kinect等深度相机，搭配Raspberry Pi-2B和Shopline自研板卡，运行自定义的实时处理框架。核心机制是外参校准：通过最小化点到平面的平方距离估计地板平面方程，获取安装高度和角度，并基于此构建背景虚拟滤波面。检测流程包括前景块提取、分层分割和基于搜索图（SG）的头部定位。系统在x64处理器上以VGA分辨率运行30 fps，在ARM处理器上为10 fps，降低分辨率可提升速度但损失精度，总准确率约95%。未来工作将扩展多相机协作、人体特征识别和遮挡轨迹拆分。\n\n### 关键要点\n- 系统采用分布式智能相机网络，每个节点包含深度相机、嵌入式计算机和通信单元，实现人员检测、跟踪与计数。\n- 深度传感器不受阳光反射或剧烈光照变化影响，但受限室内场景，可在完全黑暗中运行。\n- 相机俯视安装最小化遮挡并最大化覆盖面积，安装高度建议2.5至3.5米。\n- 外参校准通过最小化点到平面距离估计地板方程，利用深度直方图选取最显著bin加速计算。\n- 背景模型基于局部平面抬升阈值平方创建过滤面，适应非平坦地面和家具等复杂场景。\n- 人员检测分为前景块提取、分层分割和搜索图构建，利用头-肩部特征分离粘连人群。\n- 搜索图（SG）是无向图，节点代表区域，边代表连接，头部节点只有正权值连接。\n- 系统使用OpenCV、Qt和OpenNI等框架，但自研了优化的图像处理例程。\n- 计数算法采用简单的入口/出口线，通过初始区域标记和状态判断实现双向计数。\n- 实测安装高度估计误差±2 cm，在复杂场景下仍准确。\n- 在VGA分辨率下，x64平台运行30 fps，ARM平台10 fps；降低到320×280时两个平台均达30 fps但精度降低。\n- 系统整体准确率约95%，已由Shopline工业化销售。",
  "available_templates": [
    "deep_read",
    "simple"
  ]
}