Brick Masonry Surface Damage Dataset Collected from Battledore of the Darbhanga Fort (A Cultural Heritage) for Automatic Damage Detection

This dataset comprises images of superficial damages observed in the battledore of Darbhanga Fort, a cultural heritage structure located in Darbhanga, Bihar, constructed using red brick masonry. The dataset contains a total of 7,886 images, including 7,440 images with instances of ‘crack’, ‘spalling’, and ‘vegetation’, along with 446 background images that do not feature any damage. All images, with dimensions of 768 px × 768 px, are organized into two subfolders, “Damaged_images” and “Background_images”, located within the main folder “Darbhanga_Fort.”
The dataset has been utilized in the research article titled “Deep learning-based damage detection and segmentation in the battledore of Darbhanga Fort”, published in Journal of Cultural Heritage, available at DOI: 10.1016/j.culher.2025.04.023.
This dataset is a valuable resource for training and evaluating various deep learning models for damage detection and segmentation tasks.