Cultural Heritage Synthetic Depth Dataset (CHSD_rgb)

The Cultural Heritage Synthetic Depth Dataset (CHSD_rgb) is a collection of high-resolution RGB images generated from 3D models of cultural heritage objects. This dataset was created as part of a research project to evaluate the effectiveness of deep learning-based methods for monocular depth estimation. The 3D objects used in this dataset are sourced from the Virtual Hampson Museum and include 81 distinct cultural heritage artifacts.

The dataset was created using advanced 3D modeling software, and each object was captured from multiple angles using five virtual cameras positioned at -90, -45, 0, 45, and 90 degrees relative to the object. The cameras rotated around each object in 1-degree steps, producing a total of 1800 RGB images per object. In total, the dataset comprises 145,800 RGB images.