Abstract
In the field of the mixed reality technologies, the system to visualize foreground and occluded background simultaneously by semi-transparent display of foreground objects has been proposed. In this study, we propose to improve the system by using depth information captured from an RGB-D camera. The proposed system achieves robust recognition of foreground objects against illumination conditions, updates of occluded background by only a single camera, and variable transparency of objects according to the distance from the background. Experimental results show the effectiveness of the proposed method.