Figure-ground Segmentation as a Step towards Deriving Object Properties

  title={Figure-ground Segmentation as a Step towards Deriving Object Properties},
  author={Peter Nordlund and Jan-Olof Eklundh},
In this paper we discuss how gure-ground segmentation can be important in deriving object properties such as shape. Our approach is based on a conjunction of motion and depth. The main idea is to produce a 2-dimensional histogram with depth in one dimension and horizontal motion in the other. This histogram is then analyzed. The most signiicant peaks in the histogram are backprojected to the image to produce an object mask. This object mask is maintained over time. 


Publications citing this paper.
Showing 1-2 of 2 extracted citations


Publications referenced by this paper.
Showing 1-10 of 24 references

Maintenance of gure-ground segmen- tation by cue-selection," Tech. Rep. ISRN KTH/NA/P--97/05--SE, Dept. of Numerical Analysis and Computing Science, KTH, Stockholm, Sweden, Apr

  • P. Nordlund, J.-O. Eklundh
  • Int. Workshop on Visual Form, Capri,
  • 1997
Highly Influential
5 Excerpts

Data Fusion for Sensory Information Pro- cessing Systems. Dordrecht, Netherlands: Kluwer Academic Publishers

  • J. J. Clark, A. L. Yuille
  • 1990
Highly Influential
4 Excerpts

Interaction of di erent modules in depth perception,

  • H. H. B ultho, H. A. Mallot
  • in Proc. 1st International Conference on Computer…
  • 1987
Highly Influential
8 Excerpts

Active Robot Vision and Primary Ocular Processes

  • K. Pahlavan
  • Dept. of Numerical Analysis and Computing Science…
  • 1993
Highly Influential
3 Excerpts

Task-Directed Sensor Fusion and Planning

  • G. D. Hager
  • A Computational Approach. Dordrecht, Netherlands…
  • 1990
Highly Influential
7 Excerpts

Similar Papers

Loading similar papers…