A Computational Procedure for Generating Specimens of BIM and Point Cloud Data for Building Change Detection

The potential for automated construction quality inspection, construction progress tracking and post-earthquake damage assessment drives research in interpretation of remote sensing data and compilation of semantic models of buildings in different states. However, research efforts are often hampered by a lack of full-scale datasets. This is particularly the case for earthquake damage assessment research, where acquisition of scans is restricted by scarcity of access to post-earthquake sites. To solve this problem, we have developed a procedure for compiling digital specimens in both pre- and post-event states and for generating synthetic data equivalent to which would result from laser scanning in the field. The procedure is validated by comparing the physical and synthetic scans of a damaged beam. Interpretation of the beam damage from the synthetic data demonstrates the feasibility of using this procedure to replace physical specimens with digital models for experimentation and for other civil engineering applications.