SVD-based point cloud 3D stone by stone segmentation for cultural heritage structural analysis – The case of the Apollo Temple at Delphi

3D Modeling & Analysis Techniques for the Apollo Temple in Delphi

“…However, structural analysis is a critical part of asset management and can provide useful insights to improve the maintenance and restoration of these indispensable assets in an integrated and multidisciplinary fashion [4,6,7]. Future work could thoroughly compare the FEM results disclosed here against those from a fully automated mesh segmentation algorithm [44]. As mentioned, another novelty of this research paper is the ongoing multidisciplinary approach for scan-to-FEM automation, for which some promising results regarding colonnades have already been published.…”

“…Regarding finite element analysis (FEM), in particular, the major premise is that static and dynamic structural behavior of the object under investigation can be accurately modeled, thus providing facility management agents and stakeholders with a unique tool for real-time monitoring in a non-invasive, non-distractive manner [15,[39][40][41][42][43]. However, the inherent complexity of CH artifacts [1,15,40], strict legislation concerning archaeological uncertainty and standardization [43], and BIM's fundamental assumptions about regularity and homogeneity [1,40] require high-diligence cloud-to-FEM heuristics [15,39,[41][42][43][44].…”

“…Last but not least, FEM modeling entails some rigorous [2] complexity, which could jeopardize HBIM modeling in the first place [1]. Another novelty of this paper is that it was conducted with another work that targeted mesh segmentation automation for FEM applications [44]. Therefore, the results disclosed herein could provide a reference for validating the automated decimation performance under different structural analysis techniques, and, finally, recommendations and good practices for archaeology can be derived.…”

“…Three-dimensional modeling and stone-by-stone segmentation from point clouds can be performed in several ways, which have been extensively researched [1,44,53]. State-ofthe-art approaches nowadays range from semi-automating recurrent label transmission using object-oriented programming integrated with commercial BIM architectural packages [55] to deep learning semantization utilizing either well-established methods in the field of computer vision, convolutional neural networks [53,56,57], or high-efficiency voxelbased feature engineering [58].…”

“…It extends beyond the scope of this manuscript, but current literature reveals that most researchers now consolidate towards a common pipeline for converting existing CH monuments to FEM-compatible configurations [15,[39][40][41][42][43]60,61]. Scan-to-FEM workflows, with some slight variations [41,44], are initiated from the 3D texturing of the asset, also known as the digital twin [37], then render a dense point cloud [41], and either re-mesh directly for FEM or transit from BIM to FEM [32].…”