Scholarlink

<p>Thorough and systematic slope investigations are essential for preventing landslide disasters. However, current research predominantly relies on single technical approaches, hindering a comprehensive understanding of slope conditions. To address this limitation, this study proposes a novel multi-source data fusion framework that integrates unmanned aerial vehicle (UAV) optical imagery, terrestrial laser scanning, airborne light detection and ranging, and thermal infrared imaging for enhanced slope analysis. Initially, artificial embankment and landfill slopes, as well as a natural slope, were reconstructed using multi-source data and various photogrammetric techniques. Subsequently, the multiscale model to model cloud comparison (M3C2) algorithm was employed to assess discrepancies between UAV-derived point clouds and reference models. The results demonstrate that 3D slope models reconstructed from UAV optical imagery achieve centimeter-level accuracy, with close-range photogrammetry approaching millimeter-level precision. The application of the M3C2 algorithm provides a robust and quantitative means of evaluating 3D model accuracy. Additionally, the integration of multi-source data significantly improves the completeness and reliability of slope characterization, providing valuable insights for slope engineering management and disaster prevention.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

3D Slope Reconstruction and Point Cloud Analysis Using UAV Multi-Source Data

  • 期刊论文

  • 发表日期：

    2025年6月24日

查看全文
  • Dezhi Zai,
  • Jun Liu

摘要

Thorough and systematic slope investigations are essential for preventing landslide disasters. However, current research predominantly relies on single technical approaches, hindering a comprehensive understanding of slope conditions. To address this limitation, this study proposes a novel multi-source data fusion framework that integrates unmanned aerial vehicle (UAV) optical imagery, terrestrial laser scanning, airborne light detection and ranging, and thermal infrared imaging for enhanced slope analysis. Initially, artificial embankment and landfill slopes, as well as a natural slope, were reconstructed using multi-source data and various photogrammetric techniques. Subsequently, the multiscale model to model cloud comparison (M3C2) algorithm was employed to assess discrepancies between UAV-derived point clouds and reference models. The results demonstrate that 3D slope models reconstructed from UAV optical imagery achieve centimeter-level accuracy, with close-range photogrammetry approaching millimeter-level precision. The application of the M3C2 algorithm provides a robust and quantitative means of evaluating 3D model accuracy. Additionally, the integration of multi-source data significantly improves the completeness and reliability of slope characterization, providing valuable insights for slope engineering management and disaster prevention.