Abstract
In this study, we developed a landscape visualization system by using VRML (Virtual Reality Modeling Language). We developed a converter from measured data of trees to VRML format. A tree was visualized with 2 textured planes. We also made a database of two-dimensional tree images of several growth stages from 3-D plant shape generated by AMAP (a plant modeling system that produces very precise 3-D plant shape). With the developed system, landscape of Koishikawa Korakuen Gardens with 2980 trees (91 species) was visualized for validation of the system. Byusing the system, walk through simulation of the Gardens was possible in spite of the huge number of trees because one scene can be made only in few seconds. We also compared photo-realistic images of the Gardens made by AMAP system with the images made by the system developed in this study. The system can produce almost same quality images by AMAP while AMAP took about several-ten minutes to make a scene. In conclusion, the system, developed in this study, can produce a very high quality images and will be used for landscape visualization and prediction practically with very low cost.
