ABSTRACT
Camera-based bridge inspections, where one can acquire information from a distance, have gained attention. Among them, this study focused on a 3D point cloud. Point cloud data is a set of points with 3D coordinate information and is expected to obtain detailed models with the current state of structures, such as those used in FE analysis. On the other hand, bridges are constantly vibrating due to external forces such as traffic and wind. Image-based point cloud requires the acquisition of measurement target images or videos from various distances and angles to construct a 3D model. Each acquired image with a different shooting timing contains different vibration information and the effect on each image feature point may be an error factor in 3D model construction. Therefore, in this study, the accuracy of 3D point cloud construction using videos has attempted to improve by considering the effects of bridge vibration.
