We present a technique using image analysis/synthesis collaboration that, we think, could be used in the future in order to produce a 3D synthesisable representation (even partial) of a scene visualized in an image sequence.

Using computer vision techniques, we first analyze the image sequence in order to obtain a synthesisable 3D model which includes geometric, photometric and movement representations. Then we synthesize images from this model using a realistic rendering algorithm and use the difference between the real and synthetic images in order to improve the model quality and adapt the analysis/synthesis parts accordingly. An application of this technique to augmented reality will be shown

Program

Keynote Panel: Visualization of Information: Where We Are and Where Do We Go From Here

Future Directions for Information Visualization by Steven Feiner

3D Representation of Image Sequences for Future Information Highways by A. Gagalowitz

Visions, People and Notations by Bob Spence

Table of Contents