This study aims to develop a fully automatic method for more accurate and efficient deformation measurement of bridges from terrestrial laser scanning point clouds, overcoming the limitations of traditional labor-intensive inspection and manual point selection.
The proposed deformation measurement method comprises intensity-augmented Gaussian filtering to suppress environmental noise, a new fast boundary extraction algorithm for improving robustness and speed, an exhaustive-search-based corner point detection method to ensure precise cross-sectional localization and (a normal-vector-guided automatic alignment extraction algorithm to reconstruct elevation profiles.
The proposed method was validated on a multispan continuous rigid-frame bridge. The cross-sectional dimensions of four piers were reconstructed with an average root mean square error of 3 cm. The deformation of the midspan was detected to be 6 mm, satisfying the serviceability limit-state deflection limit.
The proposed exhaustive-search-based feature corner point detection algorithm outperformed the conventional Hough transform algorithm. The proposed method demonstrated both high efficiency in alignment extraction and deformation assessment, contributing to intelligent inspection of in-service bridges.
