Protected dissemination and computer-aided reconstruction of digitized cultural artifacts.

摘要

Advances in sensing technologies and data processing algorithms allow us to accurately and efficiently digitize the shape and surface characteristics of many physical objects. Among other applications, this capability presents new opportunities for preserving, sharing, and analyzing cultural heritage objects, such as artistic statuary and archaeological artifacts. This dissertation describes computational methods that we have developed and applied to allow the protected dissemination of archives of digitized 3D objects, and to aid in the reconstruction of a fragmented archaeological artifact.; The increased capability for creation and sharing of three-dimensional digital models of valuable cultural artifacts introduces new concerns for their digital rights management. We have investigated several methods of protecting 3D models from piracy and misuse, while still allowing for their interactive display and manipulation. We describe an effective approach using a remote rendering system, which protects the high-resolution 3D geometry by allowing users to access the models only via images rendered on a remote secured server. Several possible types of 3D reconstruction attacks and corresponding defenses are considered. An implementation of our protected graphics system has been employed to protect and share large archives of 3D cultural heritage objects.; Next we describe our digitization of the shape and surface appearance of the nearly 1,200 extant fragments of the Severan Marble Plan (Forma Urbis Romae), a giant 3rd-century marble map of Rome that is a primary source of topographical knowledge of the ancient city. We have created a unique digital archive of the Plan, one that includes a searchable relational database with collected archaeological information about each fragment, and that allows users to view 3D models of the fragments using our remote rendering system.; Finally, we present a set of computer matching techniques that we have applied to the problem of reassembling the fragments of the Plan. These methods utilize the geometric constraints of the fragments' incisions, shapes, and other archaeological features that we have digitized. The effectiveness of our computer-aided reconstruction approach is demonstrated with a number of new proposed fragment matches and identifications, undiscovered in prior scholarship on the Plan, that extend our understanding of ancient Rome's topography.