Method for compositing an image of a real object with a virtual scene

摘要

An improved video polygon method for combining an image of an object with a virtual scene to produce a composite image with respect to a virtual eyepoint. Two ways are provided to retain realism in the composite image. The motion of the real camera that acquires the image is coordinated with the motion of the virtual eyepoint within the virtual scene. For example, to preserve resolution as the virtual eyepoint approaches the transparent polygon on which the image is textured, the real camera is moved closer to the object. To minimize the flat appearance of the textured transparent polygon as the virtual eyepoint moves azimuthally, the transparent polygon is rotated in tandem with the azimuthal motion of the virtual eyepoint. The coordinates of the eyepoint may be provided to the computer using a 6D mouse such as a virtual camera. <IMAGE>