Archäologische Fernerkundung: Teilautomatisierte Detektion von archäologischen Befunden in hochaufgelösten Multispektral-Satellitenbildern
This interdisciplinary project aims at systematically evaluating the potential of a new generation of high resolution multispectral satellite sensors for archaeological prospection. Remotely sensed imagery can be analyzed through multispectral image analysis by semantically classifying the contents of images, and through photogrammetry by extracting geometric 3D information. Both types of information allow to determine the location and nature of exposed and near-surface archaeological residues, which is the main prerequisite for the protection and management of cultural heritage. Large-scale archaeological prospection is up to now mainly based on aerial imagery. However, currently satellite images with a spatial resolution of 1 - 0.4 m begin to rival aerial images. At the same time they are much easier to obtain, to geo-reference, and to analyze than aerial imagery. They furthermore feature an additional near infrared channel that is especially susceptible to vegetational change and thus facilitates the detection of crop marks hinting at subsurface archaeological residues. While the use of high resolution multispectral imagery for archaeological prospection has recently increased, this project is intended to extend beyond the level of previous case studies. The approach will be a systematic testing of the new data source under a variety of environmental and archaeological conditions, as well as of semi-automated methods of digital image analysis, which are standard tools in the earth sciences. The project is thus expected to contribute to the development of efficient tools for archaeological prospection.
- FB Informatik und Informationswissenschaft
|(2015): Detection of incomplete rectangular contours with application in archaeology|
Zingman, Igor; Saupe, Dietmar; Lambers, Karsten
|Exzellenzinitiative||958/08||3. Fö. Zukunftskolleg||01.10.2008 – 31.03.2010|
|Period:||01.10.2008 – 31.03.2010|