523 
PHOTOGRAMMETRIC RECONSTRUCTION AND VIRTUAL PRESENTATION OF 
MONASTERY OF CHRIST PANTEPOPTES 
Z. Duran, G. Toz 
ITU, Civil Engineering Faculty, 34469 Maslak Istanbul, Turkey - (duranza, tozg,)@itu.edu.tr 
KEY WORDS: Photogrammetry, Cultural heritage, Documentation, Reconstruction, Visualization, GPS, Modeling, Virtual reality 
ABSTRACT: 
The land of present-day Turkey, stretching out between Asia and Europe, has been called the crossroads of history. It has always 
been the scene of international exchange of culture, art and architecture. Since early days, the traditions of the past, in the social and 
cultural reflection of various Anatolian Civilizations can still be seen in Turkey and in the remains of historical cities dating from the 
Neolithic and Early Chalcolithic Ages to mosques, palaces and historical houses of the Ottoman Period. 
With are timber houses and winding streets, Fatih is a typical region of old Istanbul. As in most other areas of old Istanbul, fires have 
caused much devastation in Fatih. The monastery of Christ Pantepoptes (Eski Imaret-i Atik Cami) in Fatih was selected to sample 
building for the case study. 
The recording and documentation of the monastery of Christ Pantepoptes (Eski Imaret-i Atik Cami) has been realized by a 
combination of state-of-the-art technology: GPS measurements and tacheometry have been applied as well as photogrammetric 
techniques. 
Several methods of visualization have been applied to the data set. Besides the rendering of the photo realistic images with high 
resolution, video films have been produced and interactive inspection in a virtual reality environment has been derived. 
INTRODUCTION 
Architecture is a substantial part of our cultural heritage. But 
whereas other elements of our cultural heritage may be 
protected by putting them behind a glass in a museum, 
architectural monuments are widely used and endangered by 
long term influences like traffic or air pollution or destructive 
events causing heavy damage like earthquakes, fire or war etc. 
But by all means when monuments are seriously damaged, or 
completely destroyed, the amount and quality of any surviving 
documentation becomes highly important. 
Therefore it is necessary to document the actual state of the 
architectural monuments in a manner, which opens the 
opportunity to detect continuous damage by change detection 
techniques and to restore the monument in case of heavy 
damage. Before starting to acquire new data on the monument 
already existing data sources have to be obtained, e.g. existing 
plans of previous restorations, ancient pictures or 
documentation's of architectural research projects. 
The 3D reconstruction of buildings has been an active research 
topic in Computer Vision as well as in Digital Photogrammetry 
in recent years. Three-dimensional building models are 
increasingly necessary for urban planning, tourism, etc. (Suveg 
and Vosselman, 2000). 
Three-dimensional photo-models help us to understand spatial 
objects, even if they are not accessible for us. Using photo 
models from existing objects can make it easier for us to 
understand complex spatial structures. With the help of 
methods coming from virtual reality we can link additional 
information in the form of text or sound sequences to our photo 
models. An architect, for example, can generate an interactive 
three-dimensional information system based on a photo-model, 
which describes and visualizes an important building [Dorffner 
and Forkert, 1998]. 
Visualization efforts in Photogrammetry do not have a long 
history. In the beginning of 90’s the first photogrammetric 
attempts were mainly focused on the use of CAD software and 
the generic visualization tools they had to offer. Most of the 
research efforts at that time were concentrated to automating the 
data acquisition and processing through Digital 
Photogrammetry concepts, thus little attention was paid on how 
to add value to the data. 
2. TEXTURE MAPPING AND VISUALIZATION 
A three-dimensional photo-model is an object model where the 
texture information is taken from photographs or other optically 
working recording systems. It consists of two parts. One part is 
the three-dimensional object model in which the shape of the 
object surface is stored. Adjoining surface patches approximates 
the object itself. The second part is the photo-texture, which is 
transformed to the patches (Dorffner and Forkert, 1998). 
To visualize the derived model the photogrammetric data are 
converted to VRML (Virtual Reality Modeling Language). 
VRML is a format for 3D data with features like hierarchical 
transformations, light sources, viewpoints, geometry, animation, 
fog, material properties and texture mapping (Carey and Bell, 
1997). VRML is an open format that has become popular 
because of its suitability for publishing 3D data on the World 
Wide Web. For this reason there is a lot of software available 
that can handle VRML. This software allows a user-friendly 
interactive examination and visualization of the data. The 
conversion to VRML is fully automatic and consists of two 
parts: geometry conversion and texture mapping. In the 
geometric part the object coordinates and the topology 
information are converted. [Heuvel, 1998].