Publications / 2024 Proceedings of the 41st ISARC, Lille, France
Unmanned Aerial Vehicles (UAVs) have been widely applied for bridge inspection, and significant progress has been made to automate the UAV-enabled inspection process. However, current solutions face two primary challenges: they typically rely on pilots' experience to create a pre-defined flight path for automated image collection, and the potential for localisation errors and depth-varying bridge environments may lead to image quality issues such as blurriness, improper exposure, insufficient resolution, and coverage. To overcome these challenges, this paper introduces a novel framework that combines an objective Building Information Modelling (BIM)-based flight mission planning tool with an in-flight image-quality assessment module to ensure the comprehensive capture of high-quality images. A field experiment was carried out on a standard girder bridge to validate the effectiveness of this framework. The results show that the proposed framework: can enable the collection of bridge images via a UAV with minor human intervention and can ensure that the captured UAV images meet the inspection requirements, allowing for the reconstruction of a high-quality digital bridge model and effective detection and quantification of potential defects. The framework marks a pioneering approach for highly automated and efficient bridge inspection using UAVs and is versatile enough for application in diverse real-world bridge scenarios.