仮想物体操作を通じた視差画像からの個人立体知覚情報の獲得

概要

論文の詳細を見る
This paper discusses how to cope with failures by humans in manipulating objects in three dimensional (3D) virtual space represented by stereoscopic images with binocular disparity. Although stereoscopic images are very useful representation to visualize the depth of 3D virtual space in a computer, it is not easy for us to manipulate each virtual object precisely based on its stereoscopic images given by stereo viewing devices. One of the reasons why this difficulty occurs is that the depth recovered from stereoscopic images of 3D virtual space in the process of human visual perception is not exactly the same as the space because human visual perception is affected not only by binocular disparity but also by various kinds of factors including accomodation. As the result, manipulation of virtual objects often fails due to the difference of human perception from the virtual space in a computer. In this article, we propose to deform the virtual space so that it becomes consistent with the space perceived by each user based on the result of the user's manipulation, especially pointing the surface of each virtual object and moving that object. The experimental results show that manipulation with this modification is about four times as successful as that without modification.
社団法人映像情報メディア学会の論文
1999-06-07