CRISP SET IMPLEMENTATION ON VIDEO IMAGES FOR THE APPLICATION OF SURVEILLANCE SYSTEMS

Observing moving objects in far field’s video
surveillance is one of the main application
areas in computer vision. The strong interest in
this research direction is driven by creating full
automotive surveillance applications. This
paper presents implementing a crisp set on
video images in order to evaluate human
activities in far field’s surveillance systems.
Reducing the storage capacity in surveillance
systems is discussed also in this paper. The
concept is based on extracting two powerful
attributes from objects motion, namely velocity
and pixel frequency distribution. This step
followed by combining the measurements
mentioned above via crisp set rules in order to
evaluate the active section in the image plane
and to determine the suitable storing rate. The
experimental results proved the efficiency of the
novel approach.