"Historical Building Surface Deterioration Dataset"

"The HeritageSurfaces-23K dataset provides 23,688 labeled images that capture different surface conditions of historical buildings. The collection covers six categories: biological deterioration, chemical deterioration, cracks, human-induced damage, material loss, and undamaged surfaces. Images were collected from diverse heritage structures and manually checked to ensure consistent labeling quality.The dataset is intended to support studies in cultural heritage conservation, structural health monitoring (SHM), and computer vision\u2013based damage detection. It can be applied to a wide range of tasks, including image classification, semantic segmentation, and defect localization. The inclusion of both deteriorated and intact surfaces makes it suitable for training robust deep learning models and for benchmarking different approaches."