Information Technology

New Software Enhances Online Shopping with Realistic Images

Imagine shopping online for the perfect back-to-school outfit. You can see the colour and size and perhaps the texture of the fabric, but can you tell how it will look from different perspectives under fluorescent classroom lighting?

“The material might be very beautiful but a potential customer wouldn’t know that because the image gives a grossly incomplete sense of texture,” says Alex Vasilescu, a doctoral candidate at U of T’s Department of Computer Science. The software she has developed, called TensorTextures, requires just a few sample images of any surface with complex three-dimensional relief, such as luxurious velvets or shimmering silks. Using that information, it can render the true appearance of the surface from every possible viewpoint and under any illumination.

TensorTextures could be adapted for the Web and filmmaking and could possibly be introduced commercially within a year, says Vasilescu. The software is described in a study presented at the SIGGRAPH 2003 conference on computer graphics and interactive techniques in San Diego July 31.CONTACT: Alex Vasilescu, Department of Computer Science (currently at the Media Research Lab, New York University) 212-998-3320, maov@mrl.nyu.edu or Nicolle Wahl, U of T public affairs, 416-978-6974, nicolle.wahl@utoronto.ca

Comments (0)

Write a comment