PSI - Issue 64

Nicola Nisticò et al. / Procedia Structural Integrity 64 (2024) 2230–2237 Nicola Nisticò/ Structural Integrity Procedia 00 (2019) 000–000

2233

4

The study of paintings, aimed at detecting various pigment layers, often involves the utilization of spectroscopy, as proposed by Casini et al. (1999), and an integrated approach suggested by Venuti et al. (2020). This integrated approach includes visible, infrared (IR), and false-color infrared (FCIR) photography, X-ray radiography, investigations using X-ray fluorescence (XRF) and Raman spectroscopy, as well as Computed Tomography (CT), ultraviolet-induced fluorescence (UVF), infrared reflectography (IRR), Multi-spectral spectral imaging (MSI), hyper spectral imaging (HSI), and Time-Domain Terahertz imaging (THz-TDI). Furthermore, Scanning Laser Doppler Vibrometry (Castellini et al., 2003) can be mentioned as another technique. This method involves measuring surface velocity and enables the evaluation of structural frequencies and modes, as discussed by Scislo (2023). This approach provides an alternative to methods based on acquiring acceleration time histories through dynamic sensors and conducting Operational Modal Analysis (Reynders, 2012; Motte et al., 2015; Zahid et al., 2020). The process of object digitization and digitalization encompasses several essential steps and considerations. Segmentation and classification involve dividing the object into regions with homogeneous properties and assigning labels to each region based on its characteristics. For example, in urban modeling (Özdemir and Remondino, 2018), this can include (Fig. 1) 1) classifying regions as vegetation in order to create a masked cloud showing buildings and other Ground Level Objects (GLO) having excluded vegetations; 2) buildings segmentation. So that digitalization refers to the process of discretizing specific geometric entities based on a predefined level of detail. The chosen level of detail depends on factors such as the scale of representation, visualization, and thematic content. For instance, in the case of paintings, digitalization can be utilized for virtual restoration, which includes tasks like crack detection and removal (Spagnolo and Somma, 2010): cracks in images typically exhibit either low luminance (bright cracks) or high luminance (light cracks). The process of crack detection involves identifying these cracks within the image. When cracks are present, they may remove some of the foreground (bright) pixels from the edges of regions. Conversely, crack "closing" refers to the enlargement of boundaries of foreground (bright) regions in an image. Morphological filters, such as closing top-hat operators (CTH) and opening top-hat operators (OTH), can be employed for crack detection. CTH operators are effective in detecting dark details in an image, while OTH operators are adept at identifying bright details. Point Cloud Filtering is crucial for processing dense point clouds to create accurate 3D models, but it can be computationally intensive. Filtering techniques are employed to reduce the density of point clouds while preserving detail and accuracy. Points with poor estimated accuracy are often removed from the cloud to enhance the overall quality. The workflow for creating 3D models involves acquiring raw data and processing it to generate models with the desired resolution and accuracy. This process includes selecting appropriate measurement techniques, estimating model uncertainty, and determining the necessary level of detail.

Fig. 1. City Hall building in Dortmund: (a) original orthophoto; (b) building segmentation (Özdemir and Remondino, 2018)

The study of complex materials delves into substances with intricate microstructures and behaviors, encompassing diverse constituents with complex internal structures, distinct properties, shapes, and sizes, residing in different mediums. These materials manifest macroscopic properties like stiffness and strength, which are influenced by multiphysics processes occurring at scales beyond direct observation. Comprehending how these processes affect stiffness and strength reduction is pivotal for analyzing existing complex materials and designing enhanced ones. Developing a comprehensive modeling protocol becomes imperative for accurately predicting the thermal and mechanical properties of such complex materials.