From point cloud to parametric 3D model: polygonal modelling approach for optimising Scan to H-BIM and Scan to H-FEM processes

3D parametric modelling is a current and important research topic for the digitisation of cultural heritage in order to preserve the evolution and historical identity of an area over time; for this reason, cultural heritage assets must be subjected to continuous maintenance and monitoring, as well as valorisation and promotion. In addition, a considerable amount of these assets are often exposed to a high risk of damage caused by natural disasters and a poor state of care and neglect that leads to inevitable loss. In this context, in order to adopt all safeguarding and preservation measures in the field of CH, new sensors, digital acquisition technologies and suitable methodological approaches make it possible to create high-performance three-dimensional models, capable of increasing the detail and representation of particulars and enriching these models with heterogeneous information. However, such 3D models are often difficult to manage as they require very high processing and storage capacities. Therefore, it is necessary to define a suitable methodology that is capable of simplifying and optimising the three-dimensional models obtained from a photogrammetric or TLS approach, preserving their accuracy and metric rigour and, at the same time, performing efficiently in BIM (Building Information Modeling) and/or FEM (Finite Element Analysis) applications. The aim of this work is to identify a suitable process that is able, starting from a point cloud obtained by means of geomatic techniques, to generate three-dimensional surface models that are subsequently decimated and optimised according to the complexity of the case and the desired output. To achieve these objectives, a semi-automatic process is described that is capable of transforming a TIN (Triangulated Irregular Network) surface model into a quad-mesh polygonal model and then converting it into NURBS (Non-Uniform Rational Basis-Splines) for subsequent import into a BIM software or FEM solver for finite element calculations. As well as the optimisation requirements, this approach also pursues the objective of improving the quality of the final model, providing metrics (validated by means of suitable algorithms) and geometric resolution, resolving all the topological errors of the generated meshes (open surfaces, holes, edges, etc.), as well as optimising the computational aspects, sharing and interoperability between the different software and platforms used. The experimentation was carried out on two different case studies, and in particular a 14th-century church and a bronze statue, demonstrating the efficiency of the optimisation and management of complex geometries and validating, through a study on the repeatability of the approach considered, an efficient and accurate Scan to BIM and Scan to FEM process applied to existing cultural heritage.