Scholarlink

<p>High-profile highway-railway grade crossings (HRGCs) can pose serious safety risks due to vehicle hang-ups. Accurate 3D profiling is essential for evaluating these crossings; however, traditional methods are time-consuming, costly, and disruptive to traffic. This study evaluates the accuracy of unmanned aerial vehicle (UAV)-based techniques as an alternative for HRGC profile assessment. Field data were collected at a case study HRGC site in Oklahoma using UAV imagery and a walking profiler, with the latter serving as the ground truth reference. UAV data were processed using Structure-from-Motion (SfM) techniques to generate a digital surface model (DSM) and extract HRGC profiles. The UAV-derived profiles were compared to the ground-truth data using descriptive analysis, similarity metrics, statistical evaluations, wavelet transformation, and the International Roughness Index (IRI). Results showed strong agreement in both profile shape and magnitude, particularly within 15&#xa0;m of the railway track. Low Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) values confirmed the satisfactory accuracy of UAV-based measurements, while additional statistical indicators revealed strong relationships between UAV-derived and ground-truth profiles. The results of this study suggest that UAV-based methods offer a satisfactory, efficient, and cost-effective solution for HRGC profile evaluation, with the potential to supplement or replace traditional survey approaches.</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

Evaluation of UAV-Based Profile Measurements for Highway Rail Grade Crossings (HRGCs): A Case Study in Oklahoma

  • 期刊论文

  • 发表日期：

    2025年6月2日

查看全文
  • Kaustav Chatterjee,
  • Joshua Li,
  • Yuting Zhou,
  • Jared Schwennesen

摘要

High-profile highway-railway grade crossings (HRGCs) can pose serious safety risks due to vehicle hang-ups. Accurate 3D profiling is essential for evaluating these crossings; however, traditional methods are time-consuming, costly, and disruptive to traffic. This study evaluates the accuracy of unmanned aerial vehicle (UAV)-based techniques as an alternative for HRGC profile assessment. Field data were collected at a case study HRGC site in Oklahoma using UAV imagery and a walking profiler, with the latter serving as the ground truth reference. UAV data were processed using Structure-from-Motion (SfM) techniques to generate a digital surface model (DSM) and extract HRGC profiles. The UAV-derived profiles were compared to the ground-truth data using descriptive analysis, similarity metrics, statistical evaluations, wavelet transformation, and the International Roughness Index (IRI). Results showed strong agreement in both profile shape and magnitude, particularly within 15 m of the railway track. Low Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) values confirmed the satisfactory accuracy of UAV-based measurements, while additional statistical indicators revealed strong relationships between UAV-derived and ground-truth profiles. The results of this study suggest that UAV-based methods offer a satisfactory, efficient, and cost-effective solution for HRGC profile evaluation, with the potential to supplement or replace traditional survey approaches.