This document specifies an image-based representation model that represents target objects/environments using a set of images and optionally the underlying 3D model for accurate and efficient objects/environments representation at an arbitrary viewpoint. It is applicable to a wide range of graphic, virtual reality and mixed reality applications which require the method of representing a scene with various objects and environments.

This document:

—    defines terms for image-based representation and 3D reconstruction techniques;

—    specifies the required elements for image-based representation;

—    specifies a method of representing the real world in the virtual space based on image-based representation;

—    specifies how visible image patches can be integrated with the underlying 3D model for more accurate and rich objects/environments representation from arbitrary viewpoints;

—    specifies how the proposed model allows multi-object representation;

—    provides an XML based specification of the proposed representation model and an actual implementation example (see Annex A).