|
|
|
June 2004 |
Documentation of Sisupalgarh fortress
The scanning grid was set to a point distance of about 1 cm on the columns' surfaces. This resulted in more than 16 million 3D points. The measurements could be completed in less than 4 days. Since the scanner is very precise, no tacheometric surveys for control points were necessary. The elevation differences between some control points had to be measured with a levelling instrument since the scanner itself cannot be levelled.
The resulting point clouds were combined (registered) using 24 flat and 6 sphere type targets that had been placed at suitable locations. Farmers had planted rice in the centre of the monument; the resulting points could be edited out of the image along with some erroneous points which is unavoidable in laser scans.
Since too many points were present in overlapping scan areas, a point thinning algorithm was applied. A smoothing procedure was used before the point cloud was triangulated to produce a triangular mesh, which describes the surface. About 3 million points or 6 million triangles were needed to model the complete scene.
Finally, textures were mapped onto the surfaces. Geomagic Raindrop Studio software was used for these procedures. The virtual model formed by the triangular mesh can be visualized from any virtual observation point yielding ground plans (Fig 2), elevations or any perspective views (Fig 3 and 4). A sequence of perspectives results in an animated scene, such as a 3D flight through the site. 3D Studio Max software was used for visualization.
With these data it is possible to describe the columns in great detail, including their appearance, size and position with relation to each other. The images of the columns were by far more exact than our hand-drawings, as comparisons show. Most important thing is that we now have a site plan, and one which shows what there is and what is missing. First, trenches near the columns (Fig 2) reveal that walls have been robbed probably within the last 20 years from the west side of the site. Second, the southern part of the column complex is either disturbed or unfinished compared to the northern row.
All in all, the measuring project shows Sisupalgarh to be even more interesting than once suspected. No monument of this kind has yet come to light in western Orissa, which gives an idea of the importance of the coastal region.
Acknowledgments
A grant from the German Research Society made this study possible. Project patron was H. Kulke of the University of Kiel. We thank M. Brandtner for several years of discussions on the site with us. The authors should like to thank B.K. Rath, Director of the Orissa State Archaeology, for his support over the years.
References
|
|
|
|
|
|
|
