Camera-based Triggering of Bridge Structural Health Monitoring Systems using a Cyber-physical System Framework

Abstract

A cyber-physical framework is proposed in this paper to integrate the measured response of two highway bridges and a weigh-in-motion system located along the same highway corridor. Leveraging traffic images captured by existing traffic cameras, computer vision techniques such as histogram of oriented gradients are utilized to detect trucks in images and synchronize measured vehicle-induced bridge data and truck weights with images of vehicle loads. Those images then act as a cyber-based linkage connecting the bridge monitoring systems and a weigh-in-motion system. Such a data fusion strategy makes it possible to establish a clear input-output model for each bridge system and explore the correlation between the responses of different bridges to same trucks. In addition, the camera system can also be used to trigger bridge monitoring systems intelligently to measure more truck events with a limited monitoring system power supply.