AccessMyLibrary provides FREE access to over 30 million articles from top publications available through your library.
Create a link to this page
Copy and paste this link tag into your Web page or blog:
Techniques abound for matting objects from film or video sequences and compositing them into other scenes. Some entail matting images filmed against a blue- or greenscreen, or use manual rotoscoping to prepare the images for compositing into other settings, while newer methods transfer mattes directly from one natural background to another.
But despite advances in working with objects in a scene, relatively little progress has been made in matting and compositing the shadows cast by those objects. In fact, complex shadows typically must be created by hand or extracted from bluescreen plates and manually placed in the desired location.
Recently, researchers at the University of Washington, Microsoft Research, and Industrial Light & Magic nave unveiled a new physically based technique for extracting shadows from natural scenes and inserting them into others. The new method, which can capture simple shadows from foreground objects in one setting and alter their shape according to the backgrounds they fall onto in another, accurately conveys important visual cues about the imagery. As a result, it lends greater realism to the final scene.