A point cloud classification method for the ScanToBIM process in Archi- tectural Heritage

Heritage Building Information Modelling (H-BIM) has completely changed the meaning of manag- ing architectural sites and ancient buildings. Now- adays, the application of cutting-edge methods for analysis, conservation and restoration is made possible by the modern 3D scanning technologies, such as terrestrial laser scanners (TLS) and digital cameras, which produce highly accurate point clouds. Furthermore, as these files are time-con- suming and computationally expensive, strate- gies are being developed to optimise their han- dling and to streamline the conversion of a point cloud into a BIM model, adopting Scan-to-BIM approaches and widespread Artificial intelligence. In this scenario, the current work investigates the use of the CANUPO multiscale algorithm and the RANSAC model-fitting method for the classifica- tion of the staircase of Palazzo Nico, a neoclas- sical building in Gioia del Colle, Italy, using data obtained through TLS. The location’s geometry, which includes numerous floors, vaults, balus- trades, and typical 19th-century-style ornate staircase, makes it an excellent case study for assessing the applicability of these tools, serving as a source base for additional modelling proce- dures.