A Novel Baseline-Based Method to Detect Local Structural Changes in Masonry Walls Using Dense Terrestrial Laser Scanning Point Clouds

Abstract

The use of Terrestrial Laser Scanning (TLS) data for structural analysis is gaining growing interest. This paper is focused on a new method for detecting local structural changes in masonry walls investigated with TLS technology, which is accurate and capable to obtain dense 3D coordinates of point clouds for the measured objects. The kernel of the method is the definition of a structural change descriptor based on a baseline network. Baselines connect feature points of each brick with its adjacent bricks within one scan. To analyze changes, the baseline vectors in the baseline network of each brick are compared with the unique standard baseline network matrix, computed on those bricks without apparent damage. This paper also describes intermediate steps of the procedure, namely the TLS data acquisition, a coordinate transformation for the point cloud, and the segmentation of bricks and mortar for further extraction of individually labelled bricks. Furthermore, it illustrates the performance of the proposed baseline-based method with a validation experiment at the Stevin laboratory of Delft University of Technology. From the analysis of this experiment, interesting results were obtained: the changes along the X axis are more significant than those along the Z axis, while the changes along the Y axis only happened within the same rows of bricks of the wall. By comparing the results obtained by manually picking the brick centers, the accuracy of the proposed methodology was proved with the maximum difference value of ~4 mm.