Angkor Project Group & Scientific Computing
ArchEye - a Quadrocopter for Archaeological Applications
A complete overview map of an archaeological digging site is a valuable tool for planning reconstruction work, cataloging findings, and determining areas of risk. However, most projects don't have the financial funds or the local authorization to employ aerial observation techniques (plane or satellite) to create such maps.
In the project ArchEye we use a mechanical autonomous quadrocopter, equipped with two digital cameras to compile overview and detail maps by stitching together several individual digital photographs. By covering the complete site with a set of overlapping photos, rectifying and stitching them together, maps of high resolution can be produced in a short time without the need for a manned plane flight. The project tests a new platform for such autonomous flights and determines the practical needs of archaeologists in terms of quality, resolution and global mapping.
In a first phase, ArchEye uses a low-cost quadrocopter as the technical basis for the flight. First experiments with this platform showed the high potential of the approach as well as the wide variety of further applications that we can tackle. A more powerful platform based on an autonomous helicopter is planned for further assignments such as aerial survey for photogrammetry of temple structures and placement of independent sensors to monitor local climate changes due to deforestation and reconstruction on digging sites. For more information visit ArchEye homepage: http://pille.iwr.uni-heidelberg.de/~archeye01/kontakt.html
Global Heritage Fund
Prof. Essameddin Badreddin - Institute for Computer Engineering (ZITI), Heidelberg University
Reconstruction at the Computer: The 3-D Puzzle
In two sub-projects at the National Museum Cambodia (reconstruction of broken statues) and the Banteay Chhmar temple (reconstruction of a bas-relief wall), students from IWR and conservators working directly with these objects are trying to find a way to solve this puzzle by applying a 3-D scanner and massive computer power. The idea is to scan each of the pieces of the broken object and use mixed-integer optimization techniques to let algorithms solve the reconstruction problem.
Using this approach each piece of the puzzle has to be touched only once: for data acquisition with the scanner. An additional problem can also be tackled using this method: If several pieces of an artifact are owned by different museums, the virtual reconstruction can ascertain if these pieces really belong to the same object.
Apart from the adaption of the optimization algorithms to this new field of application, several difficult questions have to be solved to successfully complete the projects. New optimization criteria for the match between two surfaces have to be investigated. They have to be robust against different scanning techniques and the resulting differences in geometric representation of the surfaces. To improve the process they also have to take into account vector information from stone surfaces of typical Angkorian walls as well as color information taken from the surface textures - a difficult and challenging task.
National Museum of Cambodia, Phnom Penh