Date: 02 Jun 2005

Non-Euclidean object representations for calibration-free video overlay

* Final gross prices may vary according to local VAT.

Get Access

Abstract

We show that the overlay of 3D graphical objects onto live video taken by a mobile camera can be considerably simplified when the camera, the camera's environment and the graphical objects are represented in an affine frame of reference. The key feature of the approach is that it does not use any metric information about the calibration parameters of the camera, the position of the user interacting with the system, or the 3D locations and dimensions of the environment's objects. The only requirement is the ability to track across frames at least four features (points or lines) that are specified by the user at system initialization time and whose world coordinates are unknown. Our approach is based on the following observation: Given a set of four or more non-coplanar 3D points, the projection of all points in the set can be computed as a linear combination of the projections of just four of the points. We exploit this observation by (1) tracking lines and feature points at frame rate, and (2) representing graphical objects in an affine frame of reference that allows the projection of virtual objects to be computed as a linear combination of the projection of the feature points.