講演情報

[Data_Collection-25]Image-Based Monitoring and Analysis of Pavement Condition in Weigh-In-Motion Systems

*Chun-Yu Liu1, Jyun-Ping Jhan1, Eliana Chen2, Leo Chen2 (1. NTUST (Taiwan), 2. FETC International Co., Ltd)

キーワード:

Weigh-In-Motion (WIM)、Pavement Flatness Monitoring、Digital Surface Model

The Weigh-In-Motion (WIM) systems is capable of measuring the loading weight of trucks without disrupting traffic flow. However, the pavement area of the monitoring sensors may deteriorate over time due to the repeated impact of heavy traffic and high-load vehicles. Traditional monitoring methods rely on on-site inspections, which, while providing accurate results, are time-consuming, labor-intensive, and require traffic control. To comprehensively monitor the pavement condition of the WIM system, this study deployed a high-resolution multi-camera system on a highway gantry structure. Simultaneous image acquisition combined with multi-view stereo reconstruction was employed to generate long-term digital surface models (DSMs) and orthophotos of the pavement, enabling the detection of localized settlement, uplift, and progressive surface deformation through temporal differences in the derived DSMs and orthophotos. To validate the system performance, the installation height of a highway gantry was simulated by mounting the system on a building and imaging the ground from an approximate distance of 10 m. Comparison with high-resolution ground-truth data confirmed a vertical accuracy of approximately 2 mm and the ability to reliably identify surface features as small as 5 mm. Analysis of two-month monitoring data confirms that the proposed vision-based system can dynamically identify WIM pavement deterioration and provide early warnings, supporting smart highway infrastructure and intelligent pavement maintenance through continuous, high-resolution, non-intrusive monitoring.