Video frame segmentation using competitive contours

Author

Stec, Piotr ; Domanski, Marek

Author_Institution

Inst. of Control & Comput. Eng., Univ. of Zielona Gora, Zielona Gόra, Poland

fYear

2005

fDate

4-8 Sept. 2005

Firstpage

1

Lastpage

4

Abstract

The paper presents a technique for the segmentation of colour images from video sequences. The technique is aimed at the extraction of multiple objects in the presence of background motion, and without global motion compensation. A modified version of the Multi-label Fast Marching algorithm is used in the segmentation process. The most important modification allows segment merging as well as pushing back borders of other segments. Thanks to this, the limitation of a one-way propagation for fast marching is removed. The experiments described in this paper are restricted to video sequences with translational motion of rigid objects only. The side effect of the algorithm is the regularized motion field.

Keywords

feature extraction; image colour analysis; image motion analysis; image segmentation; image sequences; video signal processing; background motion; colour images; competitive contours; multilabel fast marching algorithm; multiple object extraction; one-way propagation; regularized motion field; translational motion; video frame segmentation; video sequences; Image color analysis; Image segmentation; Merging; Motion segmentation; Signal processing algorithms; Vectors;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing Conference, 2005 13th European

Conference_Location

Antalya

Print_ISBN

978-160-4238-21-1

Type

conf

Filename

7078468