A landscape database integrating photographic virtual reality

• Main Author: Zajac, Michael
• Language: en_US
• Published: 2007
• Online Access: http://hdl.handle.net/1993/1486

Methods for representing a landscape using computer databases have almost universally suffered from several shortcomings. Geographic Information Systems (GIS) and related types of mapping systems are expensive and difficult to use. They require a specialist's knowledge of cartography and mapping systems, and often need an operator to have extensive training in technology, in addition to professional knowledge in the field of study. Finally traditional GIS lacks the capability to represent the landscape visually, in a way that is useful to landscape architects and planners who must deal with visual and aesthetic issues. Recent developments in landscape visualization have been in the realm of deterministic 3D modeling. These systems are poorly suited to representing landscape at the site scale. Immersive imaging, in the form of panoramic tapestries, murals and dioramas has been used as a medium for representing visual and qualitative aspects of landscape since ancient times. Roman panoramic wall paintings, the Bayeux Tapestry and Michelangelo's Last Supper are well-known examples. At the end of the eighteenth century painted panoramas and dynamic dioramas entered the realm of technological development for popular representation of landscapes and events. Since then panoramic photography and 'omnimax' theatre have directly extended this development. Recently, computer techniques have allowed immersive imagery and photographic virtual reality (photo VR) to be produced inexpensively and used on the desktop. The tight integration of photo VR into a GIS' content and interface supports all of the characteristics of a landscape GIS, provides new information management capabilities and unprecedented accessibility. It allows the GIS to reveal an image of the landscape that has previously eluded the developers of tools for landscape design and planning.