This dataset contains 3D micro Computed Tomography (CT) reconstructions (*.tiff files) and 3D surface models (Wavefront OBJ files) of preserved impressions found on the terracotta sculpture "Study for a Hovering Putto", dated between 1735 and 1750, and attributed to Laurent Delvaux (Gent, 17 January 1696 – Nivelles, 24 February 1778, Rijksmuseum, BK-NM-9352). The preserved impressions are fingermarks and toolmarks, which we analyse in the manuscript "Artist profiling using micro-CT scanning of a Rijksmuseum terracotta sculpture" by Sero et al. (DOI: 10.1126/sciadv.adg6073).

Folder names correspond to the names assigned to each impression in the manuscript. Each folder contains a stack of *.tiff files and a "Segmentation.obj", which is the 3D model obtained from Otsu's segmentation method in Slicer3D. The stack of *tiff files and the 3D models can be visualized together in Slicer3D.

We acquired the projection images at the FleX-ray Lab (Centrum Wiskunde & Informatica), which is supported by TESCAN-XRE NV. We performed the 3D reconstructions in the same facility. We implemented 3D CT reconstructions in Python version v3.6, ASTRA v2.1.0, FleXbox v1.0.0.

We convert each grayscale intensity image into a binary image using Otsu’s segmentation method, which finds the threshold that minimizes the intra-class variance of the thresholded black and white pixels. We then convert the grayscale image into a binary one according to the computed threshold. We then load the stack of binary images in Slicer3D, where the triangular surface mesh is generated, and we export to model file (Wavefront OBJ format).

The stack of *tiff files and the 3D models can be visualized together in Slicer3D. The 3D models can be also visualized in MeshLab.