Classification of Elements in an 3D Urban Virtual Environment Using Artificial Neural Nets

Juliana Gouveia Denipote, Rodrigo Assaf

Abstract. The long time spent during the 3D modeling of the environment objects is the main problem of creating Virtual Environment cities (VE). Automatic generation tools are possible ways to increase performance when a VE is been built. This work describes an automatic generation tool using Computational Vision and Artificial Intelligence techniques to create an EV from a real city photographs.

[ project home page | contact authors ]

This project is being developed at the Laboratory of Computational Vision (LAVI) of the Engineering School of São Carlos (EESC/USP) under orientation of Profa. Dra. Maria Stela Veludo de Paiva.