Abstract (Summary)
The "virtual reconstruction of vanished heritage" is now a widespread practice around the world, due to the growing capacities of digital media to replicate and interpret lost or inaccessible cultural sites. Each effort within this growing industry has its unique technical, artistic, and interpretive challenges. This effort describes a computer-model reconstruction of The Old Main Church of Curitiba, a demolished 18th Century structure in a city in Southern Brazil. Because there was only a single reliable image of the building, the central argument of this thesis concerns the documentary and interpretive processes necessitated by this situation, involving historical research and digital reconstruction. The church demolition began in 1875, and among the remaining historical data the most reliable source of its formal features is one photograph taken in 1870. This key image was submitted to a digital rectification process based on the dimensions of a surviving clock from the church. The results provided information that allowed a virtual reconstruction and 3D visualization of the Old Main Church. Additional interpretive support came from old paintings, drawings, and maps, as well as analogies with similar contemporaneous buildings. General research on the Internet about existing practices in virtual heritage provided examples of approaches for collecting, interpreting and representing information about completely or partially vanished buildings. The analysis of these diverse cases enabled the development of a specific methodology for data elaboration using image edge detection and architectural elements simulation techniques. Currently, worldwide discourse about virtual reconstruction has shifted its main focus from technological issues and limitations to questions of data access and origin, and to decision-making within the modeling process. In particular, the production of 3D models of vanished structures introduces the problem of bias, since information must often be added to complete missing data. Usually the origin of these elements is not acknowledged. The present research, however, proposes a visualization method for classifying the data being presented, through a colored layer system that defines the levels from certainty to uncertainty in the generated model.
Bibliographical Information:


School:University of Cincinnati

School Location:USA - Ohio

Source Type:Master's Thesis

Keywords:heritage virtual reconstruction single image rectification


Date of Publication:01/01/2004

© 2009 All Rights Reserved.