Building terrain maps and virtual worlds from video imagery

This paper is the result of Cybernet efforts starting in May of 1995 and currently on-going to develop system for capturing high resolution 3D terrain and cultural features from multiple position references video sequences. The notion is to test and implement a perception-based image rendering technique for terrain capture of real sites applicable to robotics, virtual reality and simulation applications. These techniques will enable the Army to better understand and evaluate the operational capabilities of vision based task performance. The proposed development effort has focused on the development and evaluation of image-based object constructions and rendering algorithms which will be summarized following.© (1998) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.