A Semi-automatic Pipeline for the Decay Mapping and the State of Conservation Assessment of Architectural Heritage Through Point Clouds

The integration of automation and artificial intelligence has revolutionized the documentation, conservation, and restoration of architectural cultural heritage. This research presents a pipeline leveraging Machine Learning (ML) algorithms for the automatic decay mapping of digitalized architectures, combined with informative system to support experts in the assessment of a synthetic index of the conservation state. Moreover, the setup of a web-based system as a platform for the expert acknowledgment and management of decays and properties helps them in a fast and coherent final assessment. The proposed pipeline combines such technologies to the architectural recovery theories and ensures a standardized procedure compliant with international and national regulations and standards such as ICOMOS-ISCS Glossary and UNI 11182:2006, UNI 8290–1:1981 and UNI CEN/TS 17385:2019. Method and tools are applied to the Labriola Palace in Tursi, Italy, demonstrating its efficacy in assessing the conservation state of architectural heritage, as well as the higher collaborative levels to reach also among technicians with different levels of computer science skill.