Object-respecting color image segmentation

Li Hongdong; Shen Chunhua

doi:10.1109/ICIP.2007.4379141

Object-respecting color image segmentation

Li Hongdong^*, Shen Chunhua

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Citations (Scopus)

Abstract

The problem of foreground/background segmentation is of great importance in image processing and computer vision. We present a novel Linear-Programming (LP)-based algorithm for color image segmentation. This algorithm segments an image into a conceptually-meaningful foreground region (usually corresponding to the object of interest) and background regions. From a few user specified strokes we learn two Gaussian Mixture models corresponding to the foreground and background region respectively. The algorithm performs well even when the object region consists of several different colors and textures. Due to the global optimality of LP, our algorithm is free from the drawback of getting into local minima.

Original language	English
Title of host publication	2007 IEEE International Conference on Image Processing, ICIP 2007 Proceedings
Pages	II257-II260
DOIs	https://doi.org/10.1109/ICIP.2007.4379141
Publication status	Published - 2007
Event	14th IEEE International Conference on Image Processing, ICIP 2007 - San Antonio, TX, United States Duration: 16 Sept 2007 → 19 Sept 2007

Publication series

Name	Proceedings - International Conference on Image Processing, ICIP
Volume	2
ISSN (Print)	1522-4880

Conference

Conference	14th IEEE International Conference on Image Processing, ICIP 2007
Country/Territory	United States
City	San Antonio, TX
Period	16/09/07 → 19/09/07

Access to Document

10.1109/ICIP.2007.4379141

Cite this

@inproceedings{0b2a7ffbcd494f74b1fca4041c5405ab,

title = "Object-respecting color image segmentation",

abstract = "The problem of foreground/background segmentation is of great importance in image processing and computer vision. We present a novel Linear-Programming (LP)-based algorithm for color image segmentation. This algorithm segments an image into a conceptually-meaningful foreground region (usually corresponding to the object of interest) and background regions. From a few user specified strokes we learn two Gaussian Mixture models corresponding to the foreground and background region respectively. The algorithm performs well even when the object region consists of several different colors and textures. Due to the global optimality of LP, our algorithm is free from the drawback of getting into local minima.",

author = "Li Hongdong and Shen Chunhua",

year = "2007",

doi = "10.1109/ICIP.2007.4379141",

language = "English",

isbn = "1424414377",

series = "Proceedings - International Conference on Image Processing, ICIP",

pages = "II257--II260",

booktitle = "2007 IEEE International Conference on Image Processing, ICIP 2007 Proceedings",

note = "14th IEEE International Conference on Image Processing, ICIP 2007 ; Conference date: 16-09-2007 Through 19-09-2007",

}

Hongdong, L & Chunhua, S 2007, Object-respecting color image segmentation. in 2007 IEEE International Conference on Image Processing, ICIP 2007 Proceedings., 4379141, Proceedings - International Conference on Image Processing, ICIP, vol. 2, pp. II257-II260, 14th IEEE International Conference on Image Processing, ICIP 2007, San Antonio, TX, United States, 16/09/07. https://doi.org/10.1109/ICIP.2007.4379141

TY - GEN

T1 - Object-respecting color image segmentation

AU - Hongdong, Li

AU - Chunhua, Shen

PY - 2007

Y1 - 2007

N2 - The problem of foreground/background segmentation is of great importance in image processing and computer vision. We present a novel Linear-Programming (LP)-based algorithm for color image segmentation. This algorithm segments an image into a conceptually-meaningful foreground region (usually corresponding to the object of interest) and background regions. From a few user specified strokes we learn two Gaussian Mixture models corresponding to the foreground and background region respectively. The algorithm performs well even when the object region consists of several different colors and textures. Due to the global optimality of LP, our algorithm is free from the drawback of getting into local minima.

AB - The problem of foreground/background segmentation is of great importance in image processing and computer vision. We present a novel Linear-Programming (LP)-based algorithm for color image segmentation. This algorithm segments an image into a conceptually-meaningful foreground region (usually corresponding to the object of interest) and background regions. From a few user specified strokes we learn two Gaussian Mixture models corresponding to the foreground and background region respectively. The algorithm performs well even when the object region consists of several different colors and textures. Due to the global optimality of LP, our algorithm is free from the drawback of getting into local minima.

UR - http://www.scopus.com/inward/record.url?scp=48149097571&partnerID=8YFLogxK

U2 - 10.1109/ICIP.2007.4379141

DO - 10.1109/ICIP.2007.4379141

M3 - Conference contribution

SN - 1424414377

SN - 9781424414376

T3 - Proceedings - International Conference on Image Processing, ICIP

SP - II257-II260

BT - 2007 IEEE International Conference on Image Processing, ICIP 2007 Proceedings

T2 - 14th IEEE International Conference on Image Processing, ICIP 2007

Y2 - 16 September 2007 through 19 September 2007

ER -

Object-respecting color image segmentation

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this