RUF-3D Dataset

This dataset is specifically designed to provide high-quality raw data for the structural repair and analysis of building facade textures. It offers 3D models and their corresponding native, high-resolution facade textures for 320 buildings sourced from three representative Chinese cities: Shanghai, Jingzhou, and Wuhan. All data were acquired using UAV (Unmanned Aerial Vehicle) oblique photography, a technique that excels at capturing the fine details of building facades and their appearance under real-world lighting conditions. Consequently, while the dataset includes precise geometric information, its primary value lies in providing original textures that contain authentic real-world artifacts (such as occlusions, damages, and inconsistent lighting). This makes it an ideal resource for developing and evaluating a variety of algorithms for texture repair, texture enhancement, image inpainting, and material recognition. This dataset aims to provide researchers in the fields of computer graphics, computer vision, and digital twins with a benchmark platform focused on addressing real-world building appearance issues.