摄影测量学
工程类
环境科学
计算机科学
土木工程
自动化
遥感
航空摄影
法律工程学
航测
运输工程
数据处理
海洋工程
遥控水下航行器
建筑工程
数据收集
作者
Zia U. A. Zihan,Omar Ghaleb Smadi,Inya Nlenanya,Abdallah B. Al-Hamdan
标识
DOI:10.1080/10298436.2025.2579646
摘要
Unmanned aerial vehicles (UAVs) enable pavement managers to inspect larger areas more efficiently and frequently than traditional ground-based methods. However, the success of UAV-based pavement distress assessment depends heavily on photogrammetric flight parameters that influence image quality and, consequently, detection accuracy. Despite advances in automated distress extraction, limited attention has been given to understanding how flight parameters affect data quality. This study addresses that gap by evaluating the influence of UAV flight parameters on surface distress detection, focusing on crack detection, the most challenging distress to be identified from vision data. A D-optimal experimental design was employed to define an efficient parameter space using a limited number of flights over two-lane highways. Response Surface Methodology (RSM) was applied to model the relationships between flight parameters (altitude, side, and front overlap) and image quality, quantified as contrast-to-noise ratio (CNR). The resulting regression model achieved an R² of 0.78, demonstrating a reasonable fit. Using a desirability-based optimisation approach, the study identified parameter combinations that enhance image quality, providing a transferable framework for optimizing UAV-based collection of pavement distress data.
科研通智能强力驱动
Strongly Powered by AbleSci AI