Superimposing 3D Virtual Objects using Markerless Tracking

This paper presents a novel methods to estimate the coordinates of a 3D object using the four vertices of a quadrangle and to track the markerless feature points. These methods are basic and important problems in augmented reality. However, many tracking solutions are dependent on fiducial markers in video or known coordinate systems which are required to superimpose virtual objects on frames. In this paper, we begin with the fact that the rectangular objects in 3D real world are projected the perspective quadrangle onto image plane through camera. We can estimate 3D object coordinates from 4 vertices of quadrangle through transformations of image coordinates. Also, the markerless tracking method of 4 vertices is presented and the camera motion parameters between successive frames are calculated using epipolar geometry