Abstract
This study aims to construct an imaged-based Augmented Reality system. In order to archive that, estimating camera position and posture is essential in terms of alignment between real space and virtual space. In this paper, the method based on an Intelligent Space to estimate camera position and posture is proposed. Mobile camera position and posture are estimated from feature matching between that camera image and 3-dimentional feature map. 3-dimentional feature map is made up of the features extracted from camera images in the intelligent space. A concept of the proposed system and the first experimental results are shown in this paper.
