Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/9825
Title: Image segmentation by iterated region merging with localized graph cuts
Authors: Peng, B
Zhang, L 
Zhang, D 
Yang, J
Keywords: Graph cuts
Image segmentation
Region merging
Issue Date: 2011
Source: Pattern recognition, 2011, v. 44, no. 10-11, p. 2527-2538 How to cite?
Journal: Pattern Recognition 
Abstract: This paper presents an iterated region merging-based graph cuts algorithm which is a novel extension of the standard graph cuts algorithm. Graph cuts addresses segmentation in an optimization framework and finds a globally optimal solution to a wide class of energy functions. However, the extraction of objects in a complex background often requires a lot of user interaction. The proposed algorithm starts from the user labeled sub-graph and works iteratively to label the surrounding un-segmented regions. In each iteration, only the local neighboring regions to the labeled regions are involved in the optimization so that much interference from the far unknown regions can be significantly reduced. Meanwhile, the data models of the object and background are updated iteratively based on high confident labeled regions. The sub-graph requires less user guidance for segmentation and thus better results can be obtained under the same amount of user interaction. Experiments on benchmark datasets validated that our method yields much better segmentation results than the standard graph cuts and the Grabcut methods in either qualitative or quantitative evaluation.
URI: http://hdl.handle.net/10397/9825
ISSN: 0031-3203
DOI: 10.1016/j.patcog.2011.03.024
Appears in Collections:Journal/Magazine Article

Access
View full-text via PolyU eLinks SFX Query
Show full item record

SCOPUSTM   
Citations

46
Citations as of Jan 13, 2017

WEB OF SCIENCETM
Citations

27
Last Week
0
Last month
1
Citations as of Dec 24, 2016

Page view(s)

21
Last Week
0
Last month
Checked on Jan 15, 2017

Google ScholarTM

Check

Altmetric



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.