The limits of realism: perceptions of virtual landscapes

Abstract

Communicating planning results within expert groups or to local citizens is crucial to an efficient planning process. In the planning and design disciplines such as landscape and urban planning, recent digital 3D-visualizations have gained increased recognition. However, the validation of simulations of virtual landscapes, in terms of their degree of realism (R’Degree) has so far been neglected in research. This study concentrates on the question whether, how, and to which degree the real visually perceived landscape, represented through photographs, can be validly represented by means of virtual landscapes. The study area comprises the communities of Schwyz and Ingenbohl–Brunnen situated on Lake Lucerne in Central Switzerland. From a modelling point of view, landscapes are highly complex structures. Instead of manually modelling the virtual environment, which is the traditional CAD-approach, a GIS-based approach is pursued. This is the prerequisite for the efficient visualization of large data sets. The validity of the created virtual landscape is tested in an empirical study in which test persons are asked to order a set of real images and variations of the corresponding computer-generated images. In the experiment, approximately 75% of the test persons assigned the highest possible value (very high degree of realism) to one or more scenes of simulated landscapes. In order to achieve an even higher degree of realism, more and very detailed 3D-object-data and accompanying texture information would be necessary.