Point Cloud Classification and Analysis for Facade Restoration

Preserving architectural integrity floored through facade restoration is, therefore, a cross-disciplinary aspect necessary for protecting cultural and historical heritage. Moreover, the edifices are chronicles of societal growth and the passage of time and must be kept alive through time. The natural exposures to the elements and environmental pollutants, however, destroy this legacy making its preservation a necessary aspect. The concern of specialists for the restoration and maintenance of these structures has driven the application of new technologies and procedures to streamline processes. One of the initial steps in restoration is conducting exhaustive research to understand their history, architecture, and current condition. Architectural studies, conducted with cutting-edge technologies like laser scanners, are essential in this phase. These instruments can gather millions of geospatial points quickly, with millimeter-level detail, and processing these data with algorithms and software allows for the automation of many procedures. There are various techniques for processing point clouds and classification algorithms that enable a detailed evaluation of characteristics and precise planning of interventions during restoration. In our work, we explored different methods and algorithms for classifying point clouds, seeking those that provided the best results. Among them, we used the CANUPO algorithm (Automatic Classification of Point Clouds), specifically designed for the automatic classification of three-dimensional point clouds. This algorithm identifies and categorizes different groups of point clouds based on their geometric and structural characteristics, thereby facilitating the restoration and conservation process of historic buildings.