Journal of Cultural Heritage 15 (2014) 308–312 Available online at www.sciencedirect.com Case study Virtual reconstruction of paintings as a tool for research and learning Marcello Carrozzino ∗ , Chiara Evangelista 1 , Raffaello Brondi 2 , Franco Tecchia 3 , Massimo Bergamasco 4 Scuola Superiore Sant’Anna, Pisa, Italy a r t i c l e i n f o Article history: Received 9 April 2013 Accepted 27 June 2013 Available online 29 July 2013 Keywords: Virtual Reconstruction Web 3D Cultural heritage Piero della Francesca a b s t r a c t This paper presents the work related to the 3D reconstruction of the scene depicted in the famous Piero della Francesca’s fresco “The Resurrection”. The work has presented many challenges due to the fact that deliberate alterations to a mathematically correct perspective were introduced by the artist in order to visibly underline the contrast between the divine plane (Christ resurrected) and the human plane (a group of soldiers witnessing the scene). The reconstructed 3D model has been used in an interactive application enabling the virtual visit of the scene as seen from relevant viewpoints corresponding to the different perspectives and to details of the depicted figures. The application also allows to change in real-time the lighting conditions of the scene in order to compare the virtual illumination with the one present in the fresco so as to illustrate possible alternatives about the debated original collocation of the artwork. © 2013 Elsevier Masson SAS. All rights reserved. 1. Research aims The aim of this research is to provide an interactive tool to be used for educational purposes, in order to explain the use of per- spective in the painting and the alterations introduced by the artist to achieve the desired effect. A secondary aim is to investigate the lighting of the depicted scenery in order to give arguments in favour of one of the existing theories about the original collocation of the artwork as, in order to enhance the immersion of spectators, these artworks were often realized to present an illumination compatible with the one of the real hosting place. 2. Introduction and related work Virtual reconstructions of historical elements have always aroused a great interest, not only for the undoubted appeal exerted by the feeling of being transported to the past, but also for the rich opportunities offered by such tools for study and research. Exist- ing literature offers countless examples of such 3D reconstructions based on documents of several types, such as text descriptions, drawings and pictures, or on tangible facts like finds and ruins (among the others: [1,2]). Commonly, most of them share the need ∗ Corresponding author. Tel.: +39 050 882 515; fax: +39 050 883 333. E-mail addresses: m.carrozzino@sssup.it (M. Carrozzino), c.evangelista@sssup.it (C. Evangelista), r.brondi@sssup.it (R. Brondi), f.tecchia@sssup.it (F. Tecchia), bergamasco@sssup.it (M. Bergamasco). 1 Tel.: +39 050 882562; fax: +39 050 883333. 2 Tel.: +39 050 882 554; fax: +39 050 883 333. 3 Tel.: +39 050 882 553; fax: +39 050 883 333. 4 Tel.: +39 050 882 548; fax: +39 050 883 333. of balancing the attractiveness towards the general public and the scientific validity of the presented hypotheses. A peculiar case is that of virtual reconstructions of paintings. In this case, three-dimensional adaptations of an inherently bi- dimensional content offer a way to explore environments, objects and characters as conceived by artists and to observe them from points of view different from the only one originally foreseen [3]. Although every virtual reconstruction presents the risk of introducing biased filters, due to the subjective interpretation of the sources, 3D transpositions of paintings are more prone to these risks. In order to offer a wider freedom of exploration, it is required in such cases to insert elements not present in the original image because they are hidden or outside the frame of the picture. This often leads beyond the concept of interpretation, embracing the hazardous concept of “imagination”. While this kind of stretching, if aimed to a better involvement of the public, can be tolerated for recreational or popularization purposes, it is obviously not accept- able from a rigorous scientific perspective. The main purpose of such transpositions is commonly to enrich the user experience, exploiting the immersion and the interaction capabilities of virtual environments [4] in order to improve the emotional exchange and achieve a more active participation to the communication process. Two general approaches to 3D reconstructions of 2D pictorial elements exist: automatic procedures, attempting to algorithmi- cally reconstruct relevant 3D shapes referenced in the picture, and manual procedures, building on 3D modelling skills of experienced professionals. Ideally automatic techniques should be preferred because they result in cheaper and quicker processes, potentially allowing large-scale transpositions of pictures, and also because they minimize the impact of the reinterpretation introduced by the human factor. On the other side the passage from 2D to 3D 1296-2074/$ – see front matter © 2013 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.culher.2013.06.003