Low-budget 3D scanning and material estimation using PyTorch3D

Oliver Cossairt, Florian Willomitzer, Chia-Kai Yeh, Marc Walton

Abstract: In this paper, we present recent results on our effort to develop low-budget, high precision 3D scanning solutions for art museum conservators. There is a real need for this in cultural heritage applications where the potential to detect and monitor degradations is of critical importance, but financial resources for high-quality 3D scanning devices is limited to a select few larger institutions. The technology we have developed uses just a screen and camera to recover sub-mm 3D surface features in paintings, drawings, and stained-glass artworks (see Figs. 1 - 3 ). Our mobile Multi-view phase-shifting deflectometry system allows for the three-dimensional measurement of extended specular surfaces with high surface normal variations [1] . It consists only of a mobile handheld device (such as a tablet) and exploits screen and front camera for reflectance-based surface measurements.