Highly Accurate and Automated Point Cloud Registration for Full-Field Deformation Measurement and Construction Quality Inspection of a Steel-Concrete Composite Bridge Pier

Steel-concrete bridges have been widely constructed within road networks due to their distinctive mechanical properties. However, construction quality issues, such as concrete pouring blockages, delamination, and deformation, frequently arise, posing significant risks to their safe operation. Laser scanning technology provides an effective solution for rapid inspection of bridge construction quality, owing to its comprehensive information, high data density, and noncontact measurement capabilities. Nonetheless, conventional data processing methods face challenges due to their cumbersome procedures and limited accuracy in point cloud registration for deformation extraction. To address these issues, this study presents an accurate and automated full-field deformation measurement method for the construction quality inspection of steel-concrete composite bridge piers using laser scanning technology and deep learning. The innovative contributions of this study include: 1) the development of a geometric feature-based point cloud simplification algorithm that automatically selects sampling points while preserving crucial structural information and reducing data volume using the fast point feature histogram (FPFH) feature and an improved voxel downsampling technique and 2) the creation of a highly accurate point cloud registration method using geometric structure embedding and an attention mechanism to extract point cloud features, along with maximal clique constraints to estimate the positional attitude between two point cloud datasets for full-field deformation extraction and quality inspection. The effectiveness and robustness of the proposed method were validated using public point cloud datasets and field test data from a steel-concrete composite bridge pier. Results indicate that the proposed method outperforms existing feature-based registration methods in terms of translation error (TE), inlier ratio, and root-mean-square error (RMSE). Furthermore, the maximum deformations measured on the inner and outer sides of the bridge pier were 8.03 cm and 8.65 cm, respectively. The average relative error of the measured deformations was below 3% compared to the results obtained with a handheld laser scanner. Compared to existing deformation measurement methods, the proposed method achieves millimeter-level bridge deformation measurements and offers advantages in terms of noncontact, high accuracy, and automation, as it does not require sensor installation and can process point cloud data end-to-end, with minimal human intervention and low technical requirements for operators.