Learning-based classification of multispectral images for deterioration mapping of historic structures

Abstract

The conservation of historic structures requires detailed knowledge of their state of preservation. Documentation of deterioration makes it possible to identify risk factors and interpret weathering mechanisms. It is usually performed using non-destructive methods such as mapping of surface features. The automated mapping of deterioration is a direction not often explored, especially when the investigated architectural surfaces present a multitude of deterioration forms and consist of heterogeneous materials, which significantly complicates the generation of thematic decay maps. This work combines reflectance imaging and supervised segmentation, based on machine learning methods, to automatically segment deterioration patterns on multispectral image composites, using a weathered historic fortification as a case study. Several spectral band combinations and image classification techniques (regression, decision tree, and ensemble learning algorithmic implementations) are evaluated to propose an accurate approach. The automated thematic mapping facilitates the spatial and semantic description of the deterioration patterns. Furthermore, the utilization of low-cost photographic equipment and easily operable digital image processing software adds to the practicality and agility of the presented methodology.

A Rapid Pipeline for Periodic Inspection and Maintenance of Architectural Surfaces

“…Fast Random Forest has been chosen as classifier, on the basis of the results of previous research works, which identifies it as the most performing among a selection of different machine learning classifiers (Eleonora Grilli & Remondino, 2019), (Adamopoulos, 2021). Hence, the outcomes are represented by classified 2D equirectangulars/high resolution textures (Figure 6 and Figure 7), which can be visualized within a virtual immersive environment (for the 360°), for a qualitative observation of the damage distribution, or re-projected in a three-dimensional environment (for the high-resolution textures), for a quantitative measurement of their extension.…”

“…Besides, digital image processing and artificial intelligence are progressively growing, to rationalize and simplify the analysis of the collected data. Until now, machine/deep learning has been exploited, to recognize and evaluate damage patters, but mainly on 2D images, with the necessity of providing large amounts of training samples, and the lack of quantitative insights about threedimensional complex structures (Chaiyasarn et al, 2018) (Adamopoulos, 2021) (Hatir et al, 2020) (Bruno et al, 2023). In view of further simplification, easy or common tools, like spherical cameras or smartphones have been introduced, for the virtual reconstruction and representation of architectural environments.…”

Enhancing Automation of Heritage Processes: Generation of Artificial Training Datasets From Photogrammetric 3d Models

“…As a matter of fact, these procedures are often repetitive and timespending, requiring manual operations that are often unsustainable when applied in the framework of massive digitisation projects. Artificial Intelligence (AI) has established itself as a powerful and effective solution to improve the automation levels in the framework of the processing involving heritage datasets, especially concerning classification or semantic segmentation tasks (Grilli & Remondino 2019;Fan et al 2018, Zia et al 2022Adamopoulos 2021). In particular, in the framework of image processing, Convolutional neural networks (CNN) have been demonstrated to be one of the most effective -and efficient -techniques for these kinds of tasks, outperforming traditional methods in several * Corresponding author application fields (semantic segmentation, classification, object detection, etc.)…”

Predication of deterioration components of reinforced concrete columns using machine learning methods

“…Nguyen et al [48] applied Feed Forward Neural Networks (FNN) to predict maximum compressive strength. Adamopoulos, to mapping the deterioration hysterical structures is used multispectral images [49]. Avci and et al [50] diagnosed vibration based seismic damages by machine learning and deep learning methods.…”