Automated processing of large point clouds for structural health monitoring of masonry arch bridges

Abstract

Abstract Laser scanning technology is gaining popularity in a wide range of applications due to the increasing accuracy and speed at which data can be collected and an increase in laser scan data processing tools. Nevertheless, manual operations for specific applications are time consuming and can require high-performance computers to produce suitable models for further operations. Thus, laser scan data are underused in the civil engineering community. New procedures that automate the data processing for specific but repetitive infrastructure typologies are required to make full use of the technology as a basic tool for infrastructure assessment and asset management. This paper presents a new method for fully automated point cloud segmentation of masonry arch bridges. The method efficiently creates segmented, spatially related and organized point clouds, which each contain the relevant geometric data for a particular component (pier, arch, spandrel wall, etc.) of the structure. The segmentation is based in the combination of a heuristic approach and image processing tools adapted to voxel structures. The proposed methodology provides the essential processed data required for structural health monitoring of masonry arch bridges based on geometric anomalies. The method was validated using a representative sample of masonry arch bridges. The results demonstrate that this tool can provide data for further structural operations without requiring neither training in laser scanning technology nor high-performance computers for such data processing.