Work in computer vision at Irvine focuses on the development of
representations and algorithms that support the recognition of
three dimensional objects in complex scenes. Research in image
representation uses physical models for illuminants, surfaces,
and sensors to derive models that relate image observables such
as color and texture to surface and scene characteristics. These
image representations are fundamental to the development of
algorithms for image segmentation and description. An area of
particular interest is the recovery of image descriptions that
are invariant to scene properties such as the illumination
environment or the geometric arrangement of objects. These invariant
descriptions allow for efficient three dimensional recognition
in relatively unconstrained environments. Our models and
algorithms are verified and refined by experimentation with
data obtained at our image acquisition facility.
Irvine Computer Vision Laboratory
Department of Electrical and Computer Engineering
University of California
Irvine, CA 92697
phone: (949) 824-7104
fax: (949) 824-3408
Last modified: 31 January 2007
ICVL web page is maintained by:
ssarkar@uci.edu