Image-Based Models Using Crowdsourcing Strategy

Abstract

The conservation and valorization of Cultural Heritage require an extensive documentation, both in properly historic-artistic terms and regarding the physical characteristics of position, shape, color, and geometry. With the use of digital photogrammetry that make acquisition of overlapping images for 3D photo modeling and with the development of dense and accurate 3D point models, it is possible to obtain high-resolution orthoprojections of surfaces.Recent years have seen a growing interest in crowdsourcing that holds in the field of the protection and dissemination of cultural heritage, in parallel there is an increasing awareness for contributing the generation of digital models with the immense wealth of images available on the web which are useful for documentation heritage.In this way, the availability and ease the automation of SfM (Structure from Motion) algorithm enables the generation of digital models of the built heritage, which can be inserted positively in crowdsourcing processes. In fact, non-expert users can handle the technology in the process of acquisition, which today is one of the fundamental points to involve the wider public to the cultural heritage protection. To present the image based models and their derivatives that can be made from a great digital resource; the current approach is useful for the little-known heritage or not easily accessible buildings as an emblematic case study that was selected. It is the Vank Cathedral in Isfahan in Iran: the availability of accurate point clouds and reliable orthophotos are very convenient since the building of the Safavid epoch (cent. XVII-XVIII) completely frescoed with the internal surfaces, which the architecture and especially the architectural decoration reach their peak.The experimental part of the paper explores also some aspects of usability of the digital output from the image based modeling methods. The availability of orthophotos allows and facilitates the iconographic reading of the frescoes; adding to the radiometric data there is the metric potentiality of reading the proportions and composition of the organization of the frescoes. Furthermore, simplified and suitably schematized models can be even printed and can be used in a didactic environment such as the knowledge dissemination intended by the museums and other cultural institutions.Actually, this test case is even a pretext to discuss the current chance to enhance the crowdsourced collection of image data concerning the cultural heritage scattered in the world with the aim to derive spatial data and models.

Published
2016-12-22
How to Cite
SPANÒ, Antonia; HASHEMI, Narges; NOUROLLAHICHATABI, Sanaz. Image-Based Models Using Crowdsourcing Strategy. DigitCult - Scientific Journal on Digital Cultures, [S.l.], v. 1, n. 3, p. 65-79, dec. 2016. ISSN 2531-5994. Available at: <http://www.digitcult.it/index.php/dc/article/view/23>. Date accessed: 23 sep. 2018. doi: https://doi.org/10.4399/97888548993917.
Section
Articles