Three dimensional object pose estimation which employs dense depth information

摘要: The ability to accurately estimate the three-dimensional position and orientation of an object, based Solely upon Video images of the object, is of increasing interest in the field of computer vision. For example, interactive human interface applications require the ability to quickly and accurately track the pose of a user. Information regarding the user's body position must be at or near real-time, to adjust the display of the interface in a meaningful, timely manner. For instance, an application which displays a three-dimen Sional view of an object requires accurate tracking of the position and orientation of the user's head, in order to present the frames containing an image of the object from an appropriate perspective. In general, previous approaches to pose tracking often relied on assumed models of Shape, to track motion in three dimensions from intensity data, ie, image brightness. Other approaches have employed …