Color is an important and powerful cue in the distinction
and recognition of objects, however identifying or matching
the surface color of a moving object in a video surveillance
system is very critical. In this paper we describe a new approach
aiming at providing a comparison of themore widely
used color spaces in identifying colors in motion sensors.
We employ a statistical approach by conducting extensive
data mining on video clips collected under various lighting
conditions and distances from several video-cameras. The
goal is to learn how individual colors can drift in different
illumination conditions and with different color spaces.
Type:
Conference
City:
Scottsdale
Date:
2010-01-13
Department:
Digital Security
Eurecom Ref:
3027
Copyright:
© EURECOM. Personal use of this material is permitted. The definitive version of this paper was published in and is available at :
See also: