Image-based techniques for the survey of mosaics in the St Mark's Basilica in Venice

This article aims to critically examine the entire methodology of very large scale (1:1) surveying and documentation of mosaic surfaces. The term ‘survey’ should be read in its broadest and most complete and sense, including the phases of measurement and data processing as well as management and use of these data for the purposes of preservation and maintenance. The case study presented here took place at St Mark’s Basilica (Basilica di San Marco), in Venice, where mosaic flooring, wall and vault decorations have been surveyed on two separate occasions. These two experiences shared a common goal (a full-scale survey of the mosaic decorations) but differed in terms of the methodologies used, chiefly due to the technological developments of recent years. All this, therefore, lends itself to a methodological reflection and critique of the ways in which surveying technology has evolved over time. It enables to conduct surveys that would, just a few years ago, have been inconceivable due to their size and complexity. This article describes in detail current surveying processes, which includes the use of a multi-scale “image-based” approach, “re-topology” methods such as non-uniform rational B-spline (NURBS) and a tailor-made Building Information Modeling (BIM) system. This system allows the direct use of a three-dimensional (3D) model of the Basilica within the maintenance process of the monument itself with the options to georeferencing information, extract basic metric data and catalogue all its mosaics. Highlights: • Modern digital photogrammetric techniques enable the acquisition of very complex objects, not only in terms of form but also in terms of material. • To obtain high resolution orthophotos, it is necessary to accurately take care of all the stages of the process: photographic acquisition, surveying, modelling and orthographic reprojection. • High resolution images and detailed 3D models can benefit from a complex BIM system for the management of all data