Fully automated registration of 3D data to a 3D CAD model for project progress monitoring

Abstract

Abstract Alignment of the 3D data with the 3D CAD model allows for analysis of the progress of the construction project or retrieval of the desired 3D object model for use in 3D as-built modeling. The aim of this study was to propose a fully automated registration process that allows for alignment of the 3D data with the 3D CAD model. The resulting process encompasses three pre-processing steps: point cloud representation, noise filtering, and data re-sampling. Then, after the pre-processing stage, a two-step global-to-local registration procedure is applied: PCA-based global registration followed by a local registration technique that uses ICP and the Levenberg–Marquardt algorithm. The proposed process was tested through a field experiment. The experimental results demonstrate that the proposed process is not only capable of fully automating the registration of 3D data to a 3D CAD model but also beneficial for use in project progress monitoring.