Vision-based inspection of bolted joints: Field evaluation on a historical truss bridge in Vietnam

Abstract

Vision-based inspection has received significant interests from structural health monitoring and maintenance academia. The vision-based approach has unique advantages over the traditional sensor-based inspection, including non-contact sensing, low cost, simple setup, and being immune to environmental effects. Despite that, the translation of the vision-based inspection to in-service structures in Viet Nam has been limited so far. Herein, the authors examine the field applicability of a vision-based approach for joints monitoring of a historical truss bridge in Vietnam. Firstly, a well-established vision-based bolt-loosening monitoring approach is briefly described. Secondly, a field test on the Nam O bridge (Da Nang City) is performed. A digital camera is used to capture the images of representative bolted joints of the bridge. Lastly, the vision-based approach is applied to monitor the bolted joints. The angle of bolts in the joints is estimated from the captured images, from which the accuracy of the approach is evaluated. This study is one of the first case applications, demonstrating the field applicability of the vision-based bolt-loosening approach for inspecting a real bridge in Vietnam.