We will
explore the use of visual sensors in interactive systems
that augment the physical environment. The core of the course addresses
problems that relate to the digitization of attributes of human motion,
such as
the waiving of a hand or the pointing of a finger, to support natural
human
computer interaction. Elements of computational vision will be
introduced that
aim to the interpretation of visual features and their variation, into
purposeful user actions.
The
course includes an introduction
to the theory behind image acquisition and spatial interpretation of
one or
multiple images in 2 and 3 dimensions. Topics of study include
monocular,
binocular and multi-ocular active and passive vision, as well as, the
pertinent
computational method for its implementation. Practical focus will be
placed
upon the development of efficient vision modules that can be integrated
in
real-time systems, such as projector-camera interactive surfaces,
augmented
reality systems etc.
The
course will include programming labs where theoretical knowledge will
be
applied to practical assignments within a laboratory environment. In
addition,
a programming project will be assigned towards the end of the course
where the
methodologies provided in the course will be employed for the
development of a
demonstrator.