Pipeline de reconstrução 3D de artefatos culturais utilizando diferentes câmeras de profundidade e cor

Abstract

In recent years, there have been significant advances in the 3D digitization of cultural heritage and, although recent studies have sought on obtaining 3D reconstructions with a high level of detail and realism using low-cost acquisition devices, they are still limited in the context of digital cultural preservation. This is due to the challenge of reconstructing three-dimensional models with precision and quality, considering the characteristics of the data acquisition devices used. In order to obtain complete and well-detailed three-dimensional models of cultural artifacts using low-cost technology, this work proposes a 3D reconstruction pipeline that uses point clouds from different sensors, combining captures from a low-cost depth camera with high resolution RGB images acquired by an external camera processed using photogrammetry algorithms. The main contribution of this work is the description of a robust and hybrid pipeline that improves the information acquisition stage and merges data from different sensors. The pipeline was evaluated and discussed in terms of the results obtained and demonstrates its contribution to enabling the use of low-cost reconstruction methodologies in the context of digital cultural heritage. The proposed methodology was developed to be applied to the preservation of artifacts of the Federal University of Bahia's Brazilian Museum of Archaeology and Ethnology (MAE/UFBA), being also tested on table objects that enabled the validation process of the experimental results.