Fast Superpixel-Based hierarchical approach to image segmentation

Francesco Verdoja; Marco Grangetto

doi:10.1007/978-3-319-23231-7_33

Fast Superpixel-Based hierarchical approach to image segmentation

Francesco Verdoja^*
, Marco Grangetto

^*Corresponding author for this work

Not published at Aalto University

University of Turin

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

6 Citations (Scopus)

Abstract

Image segmentation is one of the core task in image processing. Traditionally such operation is performed starting from single pixels requiring a significant amount of computations. It has been shown that superpixels can be used to improve segmentation performance. In this work we propose a novel superpixel-based hierarchical approach for image segmentation that works by iteratively merging nodes of a weighted undirected graph initialized with the superpixels regions. Proper metrics to drive the regions merging are proposed and experimentally validated using the standard Berkeley Dataset. Our analysis shows that the proposed algorithm runs faster than state of the art techniques while providing accurate segmentation results both in terms of visual and objective metrics.

Original language	English
Title of host publication	Image Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings
Publisher	Springer
Pages	364-374
Number of pages	11
Volume	9279
ISBN (Print)	9783319232300
DOIs	https://doi.org/10.1007/978-3-319-23231-7_33
Publication status	Published - 2015
MoE publication type	A4 Conference publication
Event	International Conference on Image Analysis and Processing - Genoa, Italy Duration: 7 Sept 2015 → 11 Sept 2015 Conference number: 18

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9279
ISSN (Print)	03029743
ISSN (Electronic)	16113349

Conference

Conference	International Conference on Image Analysis and Processing
Abbreviated title	ICIAP
Country/Territory	Italy
City	Genoa
Period	07/09/2015 → 11/09/2015

Keywords

Bhattacharyya distance
CIEDE2000
Graph partitioning
Hierarchical clustering
Mahalanobis distance
Segmentation
Superpixels

Access to Document

10.1007/978-3-319-23231-7_33

Cite this

Verdoja, F., & Grangetto, M. (2015). Fast Superpixel-Based hierarchical approach to image segmentation. In Image Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings (Vol. 9279, pp. 364-374). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9279). Springer. https://doi.org/10.1007/978-3-319-23231-7_33

Verdoja, Francesco ; Grangetto, Marco. / Fast Superpixel-Based hierarchical approach to image segmentation. Image Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings. Vol. 9279 Springer, 2015. pp. 364-374 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{a8fde9c5fa8940baae48a80cd2c5c43b,

title = "Fast Superpixel-Based hierarchical approach to image segmentation",

abstract = "Image segmentation is one of the core task in image processing. Traditionally such operation is performed starting from single pixels requiring a significant amount of computations. It has been shown that superpixels can be used to improve segmentation performance. In this work we propose a novel superpixel-based hierarchical approach for image segmentation that works by iteratively merging nodes of a weighted undirected graph initialized with the superpixels regions. Proper metrics to drive the regions merging are proposed and experimentally validated using the standard Berkeley Dataset. Our analysis shows that the proposed algorithm runs faster than state of the art techniques while providing accurate segmentation results both in terms of visual and objective metrics.",

keywords = "Bhattacharyya distance, CIEDE2000, Graph partitioning, Hierarchical clustering, Mahalanobis distance, Segmentation, Superpixels",

author = "Francesco Verdoja and Marco Grangetto",

year = "2015",

doi = "10.1007/978-3-319-23231-7\_33",

language = "English",

isbn = "9783319232300",

volume = "9279",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "364--374",

booktitle = "Image Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings",

address = "Germany",

note = "International Conference on Image Analysis and Processing, ICIAP ; Conference date: 07-09-2015 Through 11-09-2015",

}

Verdoja, F & Grangetto, M 2015, Fast Superpixel-Based hierarchical approach to image segmentation. in Image Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings. vol. 9279, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9279, Springer, pp. 364-374, International Conference on Image Analysis and Processing, Genoa, Italy, 07/09/2015. https://doi.org/10.1007/978-3-319-23231-7_33

Fast Superpixel-Based hierarchical approach to image segmentation. / Verdoja, Francesco; Grangetto, Marco.
Image Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings. Vol. 9279 Springer, 2015. p. 364-374 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9279).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Fast Superpixel-Based hierarchical approach to image segmentation

AU - Verdoja, Francesco

AU - Grangetto, Marco

N1 - Conference code: 18

PY - 2015

Y1 - 2015

N2 - Image segmentation is one of the core task in image processing. Traditionally such operation is performed starting from single pixels requiring a significant amount of computations. It has been shown that superpixels can be used to improve segmentation performance. In this work we propose a novel superpixel-based hierarchical approach for image segmentation that works by iteratively merging nodes of a weighted undirected graph initialized with the superpixels regions. Proper metrics to drive the regions merging are proposed and experimentally validated using the standard Berkeley Dataset. Our analysis shows that the proposed algorithm runs faster than state of the art techniques while providing accurate segmentation results both in terms of visual and objective metrics.

AB - Image segmentation is one of the core task in image processing. Traditionally such operation is performed starting from single pixels requiring a significant amount of computations. It has been shown that superpixels can be used to improve segmentation performance. In this work we propose a novel superpixel-based hierarchical approach for image segmentation that works by iteratively merging nodes of a weighted undirected graph initialized with the superpixels regions. Proper metrics to drive the regions merging are proposed and experimentally validated using the standard Berkeley Dataset. Our analysis shows that the proposed algorithm runs faster than state of the art techniques while providing accurate segmentation results both in terms of visual and objective metrics.

KW - Bhattacharyya distance

KW - CIEDE2000

KW - Graph partitioning

KW - Hierarchical clustering

KW - Mahalanobis distance

KW - Segmentation

KW - Superpixels

UR - https://www.scopus.com/pages/publications/84944755004

U2 - 10.1007/978-3-319-23231-7_33

DO - 10.1007/978-3-319-23231-7_33

M3 - Conference article in proceedings

AN - SCOPUS:84944755004

SN - 9783319232300

VL - 9279

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 364

EP - 374

BT - Image Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings

PB - Springer

T2 - International Conference on Image Analysis and Processing

Y2 - 7 September 2015 through 11 September 2015

ER -

Verdoja F, Grangetto M. Fast Superpixel-Based hierarchical approach to image segmentation. In Image Analysis and Processing – ICIAP 2015 - 18th International Conference, Proceedings. Vol. 9279. Springer. 2015. p. 364-374. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-23231-7_33

Fast Superpixel-Based hierarchical approach to image segmentation

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this