{
  "paper_id": "4b6ae8d04b6b30394746318b409de812",
  "paper_title": "SLTP: A Fast Descriptor for People Detection in Depth Images",
  "paper_authors": [
    "Shiqi Yu",
    "Shengyin Wu",
    "Liang Wang"
  ],
  "publication_date": "",
  "publication_venue": "2012 IEEE Ninth International Conference on Advanced Video and Signal-Based Surveillance",
  "abstract": "This paper presents a new feature descriptor for real-time people detection in depth images. The shape cue in depth images can reduce negative impacts of variations of clothinging, lighting conditions and the complexity of backgrounds. The proposed Simplified Local Ternary Patterns (SLTP) can take advantage of depth images to describe human body shape with low computational cost. To evaluate the SLTP feature, we establish a dataset with 7260 positive samples. A series of experiments are carried out on this dataset, and the results show that the SLTP feature can achieve a high detection rate with a low false positive rate. Besides, SLTP is easy to implement, and performs fast (over 80 frames per second) on a standard desktop computer.",
  "keywords": [
    "SLTP",
    "depth images",
    "people detection",
    "feature descriptor",
    "real-time",
    "Kinect"
  ],
  "paper_institutions": [
    "School of Computer Science and Software Engineering, Shenzhen University",
    "National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences"
  ],
  "output_language": "zh",
  "provider": "doubao",
  "model": "ark-code-latest",
  "prompt_template": "simple",
  "extracted_at": "2026-08-16T19:47:34.903120Z",
  "template_tag": "simple",
  "summary": "### 摘要\n本文提出了一种用于深度图像中实时行人检测的新型特征描述符——简化局部三值模式（SLTP）。深度图像中的形状线索能够减少服装、光照条件和背景复杂度变化的负面影响。SLTP利用深度图像描述人体形状，计算成本低。为评估SLTP特征，作者建立了一个包含7260个正样本的数据集，并开展了一系列实验。结果表明，SLTP特征能以较低的误检率实现较高的检测率。此外，SLTP易于实现，在标准台式计算机上运行速度超过80帧/秒。核心机制是基于深度差值量化为三值编码，仅使用4连通像素，形成9种模式，并提取直方图作为特征。实验验证了SLTP在检测率和计算效率上优于HOG、HOD、HDD和LTP等现有特征。\n\n### 关键要点\n- SLTP是一种基于深度图像中局部深度差值的简化三值模式描述符，仅使用4连通像素。\n- SLTP编码将x和y方向的深度差值量化为+1、0、-1，形成9种模式。\n- 在深度图像中，SLTP能有效提取人体轮廓，描述人体形状。\n- 使用非重叠块（8×8像素）计算SLTP直方图作为特征描述符，无需归一化。\n- 建立SZU Kinect People Dataset，包含7260个正样本，训练集4268个，测试集2992个。\n- 负样本来自B3DO数据集，共442308个，其中训练集59508个，测试集382800个。\n- 实验确定最佳阈值$T_d=3.0$厘米，此时检测率在FPPW=10^-4时最高。\n- 选择48×96像素的检测窗口在计算成本和检测率之间取得较好平衡。\n- 与HOG、HOD、HDD和LTP相比，SLTP在深度图像上获得更高的检测率。\n- SLTP特征提取仅包含减法、逻辑运算和模式统计，无浮点运算，计算效率高。\n- 在320×240图像上，SLTP耗时11.6毫秒（86.2 FPS），比HOG快约7倍。\n- SLTP直方图维度仅为LTP的9/118约1/13，进一步降低了计算量。",
  "available_templates": [
    "deep_read",
    "simple"
  ]
}