Populate’s AvatarBooth is working

Will Populate’s AvatarBooth fly? asked Peter Champ and Stephen Crampton in the first issue of i3magazine. We now know that it will and does: Populate’s automated AvatarBooth is up and running.

Avatars are the electronic representations of human beings in virtual environments. They can take any shape or form – from a simple icon to a fully-rendered, photorealistic representation of the human form.

Avatars are generated in the AvatarBooth. This is like the ubiquitous Photo Booth you see at railway stations and elsewhere, except that the image created is 3-dimensional instead of 2-dimensional.

The generation of avatars is based on a technique pioneered by Adrian Hilton and his team at Surrey University. In each pose, two photographs are taken: a backlit photograph which creates a silhouette image ; and a flash-lit colour image. The silhouettes from the four backlit photographs are used to calculate four outlines of the person. These outlines are compared with similar outlines for a generic avatar, and it is calculated how the standard avatar must be distorted to fit the person’s outlines. These distortions are then applied to the three-dimensional model of the generic avatar, to create an avatar with the person’s individual shape. Finally, parts of the colour photographs (textures) are pasted onto the distorted avatar. The result is a personalised, photo-realistic avatar.

Web: www.3dscanners.com/populate/home.htm

home + about i3net + services and publications + CI projects + ESE projects
esprit + european commission + IST

© 1998,1999,2000 i3net -- Search i3net -- About the i3net web pages
Revised: 01 July, 1999. Mail to webmaster@i3net.org