ISSN : 1796-2048
Volume : 1    Issue : 1    Date : April 2006

Human Tracking by Fast Mean Shift Mode Seeking
C. Beleznai, B. Frühstück and H. Bischof
Page(s): 1-8
Full Text:
PDF (675 KB)

Change detection by background subtraction is a common approach to detect moving foreground.
The resulting difference image is usually thresholded to obtain objects based on pixel
connectedness and resulting blob objects are subsequently tracked. This paper proposes a
detection approach not requiring the binarization of the difference image. Local density maxima in
the difference image - usually representing moving objects - are outlined by a fast non-parametric
mean shift clustering procedure. Object tracking is carried out by updating and propagating cluster
parameters over time using the mode seeking property of the mean shift procedure. For occluding
targets, a fast procedure determining the object configuration maximizing image likelihood is
presented. Detection and tracking results are demonstrated for a crowded scene and evaluation of
the proposed tracking framework is presented.

Index Terms
automated visual surveillance, motion detection, mean shift clustering, human tracking, occlusion