In the last years, Evolutionary algorithms as well as other bio-inspired approaches
have been shown to be tools which can be used effectively in the development of
systems (software or hardware) for Computer Vision, Graphics and Robotics techniques.
We strive to provide a high quality special issue on behalf of the Pattern Recognition Community, 
which illustrates the state-of-the-art in Evolutionary Computer Vision and Image Understanding.

We are soliciting papers on these topics to appear in a special issue of Pattern Recognition
Letters. Papers appropriate to this issue will have theoretical and experimental evaluation
of vision algorithms as their central focus, with the application cutting across the following
areas, but which are not necessarily limited to:

1) Low-level vision

Evolutionary optimization, adaptation and learning algorithms for edge detection,
image segmentation, figure-ground separation, texture analysis, feature selection,
shape-from-shading, denoising and surface reconstruction.

2) High-level vision

Evolutionary computation for object recognition, scene analysis, indexing and search
of model/image databases, and high-level learning of symbolic descriptions.

3) Active vision

Evolutionary computation for autonomous vision-guided navigation, path planning, sensing
strategies, sensor integration, visual servoing, vergence and gaze control, hand-eye
coordination, active tracking and vision-guided task planning.

4) Neural vision

Evolutionary computation for learning, adaptation and optimization of neural network
structure and topology for computer vision problems.

All submitted papers will be reviewed according to the guidelines and standards of Pattern Recognition Letters.
All submissions will be peer reviewed for originality, significance, technical content and relevance to the special issue.
A digital manuscript, as well as, 5 hard-copies (not exceeding 20 double-spaced pages in 12 pt size font including
figures and tables) describing previously unpublished and original research should be submitted to reach the guest
editor on or before October 31, 2004.

Dr. Gustavo Olague
Computer Science Department
Applied Physics Division
CICESE, Research Center
P.O. Box 434944
San Diego, CA., 92143-4944
USA
e-mail: olague@cicese.mx