GISdevelopment.net --> Application --> Archaeology
The recording of Bet Giorgis, a 12th century Rock-Hewn church in Ethiopia
Dept. of Geomatics
University of Cape Town, South Africa
Dept. of Geomatics
University of Melbourne, Australia
Armin Gruen & Thomas Buehrer
Institute of Geodesy & Photogrammetry
Around 1200AD, King Lalibela, one of the last kings of the Zagwe dynasty that ruled in Northern Ethiopia for 200 years, created a number of remarkable rock churches which remain to this day as functional places of worship. They also constitute symbols of paramount religious, cultural and architectural significance to the people of Ethiopia. Of the basically three types of these rock churches, one appears to stand apart in its uniqueness. This is the rock-hewn monolithic church, which while imitating a built-up structure is actually cut in one piece from the rock and separated from it by an all-around trench. The best known of the monolithic churches is Bet Giorgis (St George's Church) in the town of Lalibela. Given the national and international significance of this World Heritage site, a project was recently initiated to undertake a photogrammetric documentation of Bet Giorgis. This project is expected to culminate in the production of a fine-detail visually realistic digital model of the church and its immediate surroundings. The work has entailed initial image recording and photogrammetric triangulation to produce a numerical frame of reference (completed in October, 2000), and attention is now being paid to line and surface model generation, orthorectification, texture mapping and visualisation. This paper reports on the aims of the Bet Giorgis project and discusses progress made to date.
The town of Lalibela lies in the province of Wollo in northern Ethiopia, some 640 km from Addis Ababa. Other than on market day, Lalibela is not much more than a quiet mountain village, yet it is internationally renowned for its rock-hewn churches. The creation of these churches is ascribed to King Lalibela, one of the last kings of the Zagwe dynasty. All 12 churches in the town are thought to have been constructed within a 100-year period around 1200 AD. The denomination of the still functioning churches is Ethiopian Orthodox, which is part of the Coptic Christian Church headquartered in Alexandria.
Of the three basic types of rock-churches in Ethiopia: built-up cave churches, rock-hewn cave churches and rock-hewn monolithic churches, the last, which is of current interest, is unique to the Lalibela region. Arguably the most significant of Lalibela's four strictly monolithic rock-hewn churches is Bet Giorgis (the Church of Saint George), which is regarded as being the most elegant and refined in its architecture and stonemasonary. Figure 1 shows the 12 x 12 x 13m Bet Giorgis standing within its 25m square trench.
Figure 1: Bet Giorgis rock-hewn church, Lalibela.
Legend has it that Bet Giorgis, which stands apart from the two main groups of rock churches in Lalibela, was built only after King Lalibela was reproached by Saint George (the national saint of Ethiopia) for not having built a house for him. King Lalibela's response was to build a church, the construction of which, legend tells, was supervised by Saint George in person. As is apparent from Figure 1, the 'construction' of a monolithic rock church was in fact an excavation, the procedure being to first cut free a block of stone in the volcanic tuff, after which stonemasons chiselled out the church, shaping both the exterior and interior. The extent of the detail involved in this process can be appreciated from Figure 1, as well as from the cut-away sketch plan of Bet Giorgis (Figure 2).
Figure 2: Interior layout of Bet Giorgis.
Bet Giorgis is positioned in its deep pit on a sloping rock terrace with the church being accessed via an entrance trench and tunnel. Around the walls of the courtyard in the pit there are caves and chambers which house both today's priests and the graves of pious former pilgrims and monks. The cruciform church rises approximately 12m from its triple-stepped supporting platform, and it has three west-facing doorways (characteristic of Ethiopian churches), nine 'blind' lower level windows and 12 upper-row windows with semi-palmette cross motifs. The interior of the church follows the cruciform floor plan (Figure 2) and on the roof there is a relief of three equilateral Greek crosses inside each other (Figure 1).
The roots of Lalibela rock churches are thought to lie in Axumite architecture and in the early Christian basilica, yet while they may reflect a blending of eastern Mediterranean Christianity and Axumite tradition, they are also a truly unique contribution to Ethiopian Christian heritage. In recognition of their significance they have been accorded UNESCO World Heritage status.
As an aid in the long-term preservation of Bet Giorgis, and as a contribution towards making this remarkable heritage site more accessible in today's 'virtual world', a project was undertaken to photogrammetrically document the church. The project, which was initiated by the first author, with support of government agencies in Addis Ababa and encouragement from UN affiliated agencies, has as its ultimate aim the creation of fine-detail visually realistic digital models of both Bet Giorgis and other Lalibela rock churches. With direct support being given to the project by the Ethiopian Mapping Authority, work commenced in October, 2000 with a field trip to Lalibela. All necessary imagery for the photogrammetric reconstruction was recorded during this time, along with the gathering of supplementary data necessary for a comprehensive mapping of the Bet Giorgis site.
Three fundamental data sources were involved in the Bet Giorgis documentation. The first of these was existing 1:10,000 scale aerial photography from which ortho-imagery and a DTM of the Lalibela area and church surroundings could be generated. The second was ground survey data to facilitate control for the aerial imagery and to provide a tie between the local XYZ reference coordinate system used in the close-range photogrammetry and the geodetic network. The third was imagery, to facilitate photogrammetric triangulation, ortho-image generation and texture mapping for the visually realistic Bet Giorgis model. In the present account, the discussion will be confined to the close-range photogrammeric survey.
In recognition of the flexibility and favourable metric performance of modern low-cost, off-the-shelf digital cameras, a decision was made to accomplish the 3D measurement of significant feature points on the church and pit walls by photogrammetric means, and to fully carry out this task on site. As a consequence it was planned that upon leaving Lalibela after the 4-day fieldwork phase, a sufficiently dense array of feature points would be surveyed. These would provide a preliminary point cloud for the later to be determined CAD model of the church, and facilitate straightforward exterior orientation determination for any images that were to be employed in subsequent ortho-image generation and model rendering. Thus, facilities were required in the field to support camera calibration and photogrammetric bundle adjustment, as well as supplementary object point triangulation (with given exterior orientation) and exterior orientation determination. These needs were all met by the Australis software system for off-line digital close-range photogrammetry (Fraser and Edmundson, 2000).
A further consequence of the use of 'amateur' CCD cameras as metric imaging devices is that on a project such as the Bet Giorgis modelling there can be expected to be no shortage of cameras. In this instance, comprehensive sets of imagery were recorded with three digital cameras, a Kodak DCS330, a DC210 and a SONY Cybershot. Also, a Leica R5 35mm semi-metric analogue camera was employed for comparative studies. The focus of the SONY Cybershot was manually set to infinity and the zoom to the shortest focal setting, and calibration parameters from a prior laboratory testfield calibration were employed. The DCS330 was calibrated for two lenses (focal lengths of 14mm and 28mm) in the field via self-calibration, as was the DC210. This highly automated process consumed only a few 10s of minutes for each camera/lens combination and required the establishment of an array of 40 retrotargets on the wall of a nearby building. Confirmation of this stand-alone calibration was also afforded in the subsequent multi-image triangulation used to position interest points on the church.
These 'interest points' comprised both artificial targets attached to the rock face and natural feature points on the church. The artificial targets, 2.5cm white dots on a black background, would constitute 'framework points' for an optimal accuracy, multi-image bundle adjustment. The natural points, on the other hand, tended to be located on features of interest for the building of a CAD model, e.g. on edges, corners, etc. Moreover, within Australis the artificial targets could be measured automatically to, nominally, 0.1 pixel accuracy, whereas the natural points required manual image measurement to about 0.5 - 1 pixel accuracy. Shown in Figure 3 is a representative DC210 image highlighting targeted (green) and natural points (yellow). Beyond a height of about 6m it was necessary to use natural points due to lack of access for affixing targets.
Figure 3: Image measurement within Australis.
Figure 4 illustrates the network of 57 DC210 images and 210 object points which was computed within Australis to provide an initial array of accurately positioned object points for subsequent exterior orientation determination (for images not in the network) and point densification through spatial intersection. Redundant, surveyed control points were employed to transform the reference coordinate system of the photogrammetric survey into the geodetic system. An analysis of checkpoint residuals indicated a photogrammetric triangulation accuracy of close to 1.5cm, which was consistent with expectations for an average imaging scale of 1:3000 (4.8mm lens and 15m object distance) and the 1-pixel measurement precision anticipated for the natural targets.
As is indicated in Figure 4, images were recorded at intervals of a few metres around the top of the pit and also around the bottom, where stations were primarily at the front and rear of the church. The figure shows only the images employed to establish the framework points and not those taken as stereopairs for purposes of fine-detail feature modelling and texture mapping. The photogrammetric triangulation phase was not restricted to use of the DC210. Similar networks were also processed for the DCS330 and Sony Cybershot imagery using Australis and Photomodeler (www.photomodeler.com), respectively, with consistent results being obtained. From the full block of SONY Cybershot images, only those of the church itself have been used so far for triangulation and CAD model generation (40 images recorded from the top of the trench and 31 from the bottom). The reason for carrying out a separate triangulation with these images, based on the points generated by the DC210 block, stems from a particularity of Photomodeler. To build up a CAD model with Photomodeler it is more convenient to include all images in a single project, since this affords faster access. Also, the resulting CAD model can be expected to be more geometrically consistent and free of problems in the overlap regions of photogrammetric models. Moreover, experience has demonstrated that if only single models are measured sequentially, points and features in the overlap areas are very often either missed or measured twice.
Figure 4: 57-station, 210 point photogrammetric network to provide 'framework' points.
The result of the photogrammetric triangulation phase is a point-cloud of feature points, some of which are useful to the CAD model/wireframe generation process, and some which while not being of prime use in this regard are nevertheless crucial to the later processes of image registration, ortho-image generation and texture mapping. The next requirement, however, is generally to densify the object point array in order to include feature points needed for the geometric modelling. Given that there was already a comprehensive network of images of known exterior orientation, this task involved straightforward spatial intersection from typically just 2-4 images, a process which can be carried out interactively and rapidly in either Australis or Photomodeler. The majority of such points were measured in individual stereomodels via monoscopic image measurement. It is noteworthy that every new feature point was also included in the bundle adjustment such that the strength of the network improved in a stepwise fashion with the collection of CAD model points.
Visual interpretation and manual measurements were found to be absolutely necessary for producing a decent CAD model, and at this point there was little room for automation. The church possesses a much larger amount of surface detail than was expected at first sight, and the primary surfaces contain many irregularities in geometry, such that straight edges of substantial length and planar surfaces of significant size are not in abundance (e.g. features include horizontal mouldings, projecting corner beams in the doors, projecting gutter spouts and crowned ogee-arches in the upper row of the windows). Therefore, a hierarchical measurement strategy was used with feature refinements at each iteration step. The resulting point cloud was then turned into a line model and finally into a surface model. Since these two operations were carried out manually, they constituted the most time consuming of the reconstruction processes.
In order to realise the goal of a fine-detail visually realistic model of Bet Giorgis, a fully rendered solid model is required. Two approaches were adopted for this phase, one at UCT-Cape Town and the other at ETH-Zurich. The former essentially comprised the following steps: exterior orientation determination, image rectification/ortho-rectification, image registration and texture mapping. The latter employed a self-developed, view-dependent texture mapping procedure, as described in more detail in Visnovcova et al. (2001). In this object-oriented technique, the texture of each object face is mapped from the original images directly, with a selection of the geometrically optimal source image being included. Also, in order to avoid checkerboard-type texture artifacts, mapping can be from multiple images with the texture elements being combined from several adjacent images in a weighted average procedure. The visualisation program disp3D imports object faces for texture mapping from Photomodeler. At this writing (January, 2001), texture mapping is in the process of being completed. Shown in Figure 5 is a very crude preliminary textured model of Bet Giorgis generated by 'rubbersheeting' imagery onto the implicitly assumed planar walls of the church. This model, when scrutinised closely, highlights the deficiency of rubbersheeting techniques for accurate solid model renderings of structure with significant relief.
Figure 5: Screen capture from preliminary VRML model created by 'rubbersheeting' of imagery.
A primary aim of building a visually realistic digital model is to make it available to as wide a group of interested parties as possible. This naturally implies dissemination of the model in an international standard format that support 3D interactive viewing, for example VRML for web delivery. As an example of the functionality of visualisation software suited to this task, the program disp3D, which uses C, Motif and OpenGL graphics libraries, supports DXF, VRML, V3D (an internal format, used by CyberCity Modeler), TIFF, JPEG and PPM for input; and DXF, VRML, TIFF and JPEG for output. This means that most of the relevant commercial visualisation packages can be used with the generated Bet Giorgis model. Disp3D delivers the following visualisation functions: point cloud; wireframe with hidden lines; shaded surface; and texture map in mono and stereo display, via anaglyph projection and polarized stereo. Image streams for MPEG-1/MPEG-2 videos can also be created, and a fly-over and walkthrough path definition capability is currently under development. These features, as applied to the Bet Giorgis model, will be shown at the time of this paper's presentation. The model will also be made available in the near future on the Web.
Patias & Peipe (2000) have spoken of an ever-changing paradigm in the photogrammetric recording and documentation of cultural heritage sites. They have highlighted recent enhancements in the photogrammetric approach. These include a trend away from stereomodels to multi-image surface reconstruction; the use of low-cost, amateur digital cameras and user-oriented image measurement and data processing systems; the potential for multi-sensor integration; the widening acceptance of image-enhanced information systems; and the widespread use of 3D modelling, visualisation, web-authoring tools and information delivery via the Internet.
The aims and conduct of the Bet Giorgis documentation are very much in accordance with this new paradigm, and it is the hope of the authors that once the visually realistic digital modelling of this church and others in Lalibela are complete, a valuable information source will have been generated for use in site conservation, historical and architectural studies, education related to Ethiopia's rich heritage and history, and in visualisation for general enquiry and tourism.
The authors gratefully acknowledge the support and assistance of the Ethiopian Mapping Authority in the conduct of the Lalibela Project. The assistance of Mr Hadgu Medhin and the encouragement of Mr Tsegaye Denboba Wolde from the EMA have been much appreciated. The authors also acknowledge assistance given in the fieldwork and data processing stages by M.C. Biers and Alemu Nebebe. Li Zhang is thanked for providing his texture mapping and visualisation program disp3D.
|© GISdevelopment.net. All rights reserved.|