Ebert & Associates' research to explore the development of methods for the 3-dimensional recording of rock art, and of integrating such data within a unified database for analytical and management uses was supported by National Science Foundation Phase I and Phase II Small Business Innovation Research grants. A major purpose of our research has been to remedy the lack of coherence in rock art recording by developing a recording method that is effective, efficient, makes the best use of existing hardware and software, and collects and organizes data in a format that will be accessible to the widest number of investigators.
Petroglyphs in the SBIR Project sample were photographed using both 35 mm and digital cameras. Prior to photography, targets for use as total station control points were positioned on the rock panel. Photographic and topographic recording of rock art panels using the digital camera and the reflectorless total station were conducted sequentially so that the targets used to register the digital image to its topographic background were recorded in their exact placement in each medium.		Extraction of the glyph from the georeferenced raster image is based on selections of color values in the raster's attribute table that best represent the shape of the glyph. Cells selected from the raster surface are converted to vector data in the form of a shapefile. This is a subtractive process rather than an additive process, that eliminates many of the decisions about what to include as part of the design.	The georeferenced shapefile created can be viewed in a GIS environment with other data layers.
The petroglyph classification algorithm can identify subtle features which hand drawn recording can sometimes overlook, such as the headdress on this petroglyph figure.
Sketch of petroglyph panel elements and targets (on left) and automated petroglyph extraction results of selected glyphs (on right).
© 2003 Ebert & Associates, Incorporated