Abstract
The practical usefulness of many important automated
systems is currently dependent on the advancement
of computer vision to a point where well defined
engineering and computer science methods can
be used in the development of practical equipment.
In this proposal I will suggest an approach to developing
practical robust computer vision systems. I
propose to develop an intelligent architecture, in collaboration
with research partners, which will permit
the development of algorithms for implementation in
specific engineering development projects as well as
the expansion and validation of the theoretical understanding
of computer and human vision and intelligence.
In this paper I will describe a new approach
to developing computer vision systems which will include
the development of an overall system based on
three parts. First we will study aspects of human and
computer vision and intelligence. The second part of
the project will be the development of new AI algorithms
for application to specific problems and issues
in computer vision. The last part of the project will
be the integration of these new algorithms and existing
techniques under a new architecture so that
more complex, less restricted applications can be address.
These parts of the research and development
effort should be worked at the same time and by the
same team to maximize collaboration.