Scholarlink

<p>3D laser scanning has been widely used for digital modeling of large-scale structures, yet long-span bridge reconstruction remains challenging due to the limitations of single-source data acquisition. Terrestrial laser scanning (TLS) provides high-precision data but suffers from incomplete coverage due to scan range and occlusion issues. In contrast, airborne laser scanning (ALS) can capture full bridge geometry but struggle with data noise. This study proposes a heterogeneous point clouds fusion method that integrates TLS and ALS data to reconstruct long-span bridge’s point cloud models with high accuracy. The method aligns these two data sources using stable structural references (e.g., bridge towers), achieving a low registration error. Based on the fused model, we develop a multi-stage cable shape recognition method to accurately identify main cable morphology. Field validation on a Yangtze River bridge demonstrates significant performance gains. Compared with single-source methods, the proposed fusion approach improves main cable completeness by 161.94% over TLS and 6.98% over ALS, and enhances shape continuity by 17.90% and 71.93%, respectively. This framework offers a robust and efficient solution for bridge modeling and structural assessment, supporting downstream applications such as health monitoring and deformation analysis.</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

Heterogeneous point clouds fusion method for long-span suspension bridge cable morphology analysis

  • 期刊论文

  • 发表日期：

    2026年5月26日

查看全文
  • Yanjie Zhu,
  • Ziyang Tang,
  • Wen Xiong,
  • Luyao Sun,
  • Muxuan Li

摘要

3D laser scanning has been widely used for digital modeling of large-scale structures, yet long-span bridge reconstruction remains challenging due to the limitations of single-source data acquisition. Terrestrial laser scanning (TLS) provides high-precision data but suffers from incomplete coverage due to scan range and occlusion issues. In contrast, airborne laser scanning (ALS) can capture full bridge geometry but struggle with data noise. This study proposes a heterogeneous point clouds fusion method that integrates TLS and ALS data to reconstruct long-span bridge’s point cloud models with high accuracy. The method aligns these two data sources using stable structural references (e.g., bridge towers), achieving a low registration error. Based on the fused model, we develop a multi-stage cable shape recognition method to accurately identify main cable morphology. Field validation on a Yangtze River bridge demonstrates significant performance gains. Compared with single-source methods, the proposed fusion approach improves main cable completeness by 161.94% over TLS and 6.98% over ALS, and enhances shape continuity by 17.90% and 71.93%, respectively. This framework offers a robust and efficient solution for bridge modeling and structural assessment, supporting downstream applications such as health monitoring and deformation analysis.