PhD Qualifying Examination


"Graph Based Image Segmentation"

Mr. Jingdong Wang


Abstract:

Visual grouping, as the basis of image understanding and
recognition, has been an active research area in computer vision.
The past twenty years have seen the large progress on it.
Recently, graph based approaches have been demonstrated to be the
state of the art. In this survey, we review two approaches. One is
called spectral graph partitioning, which has obtained large
success on automatic image segmentation. The other is
maxflow/mincut based combinational optimization approach. It has
been validated on interactive image segmentation. Many
experimental results are shown in this survey. Finally this survey
is concluded by possible future work.


Date:     		Monday, 6 March 2006

Time:                   2:00p.m.-4:00p.m.

Venue:                  Room 2404
			lifts 17-18

Committee Members:      Prof. Long Quan (Supervisor)
			Dr. Huamin Qu (Chairperson)
			Dr. Chiew Lan Tai
			Dr. Dit Yan Yeung


**** ALL are Welcome ****