Dense Iterative Contextual Pixel Classification using Kriging

Melanie Ganz; Marco Loog; Sami Brandt; Mads Nielsen

doi:10.1109/CVPR.2009.5204055

Dense Iterative Contextual Pixel Classification using Kriging

Melanie Ganz, Marco Loog, Sami Brandt, Mads Nielsen

1 Citation (Scopus)

Abstract

In medical applications, segmentation has become an ever more important task. One of the competitive schemes to
perform such segmentation is by means of pixel classification. Simple pixel-based classification schemes can be improved by incorporating contextual label information. Various methods have been proposed to this end, e.g., iterative
contextual pixel classification, iterated conditional modes, and other approaches related to Markov random fields. A
problem of these methods, however, is their computational complexity, especially when dealing with high-resolution
images in which relatively long range interactions may play a role. We propose a new method based on Kriging that
makes it possible to include such long range interactions, while keeping the computations manageable when dealing
with large medical images.

Original language	English
Title of host publication	IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 2009. CVPR Workshops 2009.
Number of pages	7
Publication date	2009
Pages	87-93
ISBN (Print)	978-1-4244-3994-2
DOIs	https://doi.org/10.1109/CVPR.2009.5204055
Publication status	Published - 2009
Event	IEEE Computer Society Workshop on Mathematical Methods in Biomedical Image Analysis (MMBIA 2009) - Miami Beach, United States Duration: 20 Jun 2009 → 25 Jun 2009 Conference number: 10

Conference

Conference	IEEE Computer Society Workshop on Mathematical Methods in Biomedical Image Analysis (MMBIA 2009)
Number	10
Country/Territory	United States
City	Miami Beach
Period	20/06/2009 → 25/06/2009

Access to Document

10.1109/CVPR.2009.5204055

MMBIA finalv4Final published version, 302 KB

Cite this

Ganz, M, Loog, M, Brandt, S & Nielsen, M 2009, Dense Iterative Contextual Pixel Classification using Kriging. in IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 2009. CVPR Workshops 2009. . pp. 87-93, IEEE Computer Society Workshop on Mathematical Methods in Biomedical Image Analysis (MMBIA 2009), Miami Beach, United States, 20/06/2009. https://doi.org/10.1109/CVPR.2009.5204055

@inproceedings{ba5aa4e0a2b011debc73000ea68e967b,

title = "Dense Iterative Contextual Pixel Classification using Kriging",

abstract = "In medical applications, segmentation has become an ever more important task. One of the competitive schemes toperform such segmentation is by means of pixel classification. Simple pixel-based classification schemes can be improved by incorporating contextual label information. Various methods have been proposed to this end, e.g., iterativecontextual pixel classification, iterated conditional modes, and other approaches related to Markov random fields. Aproblem of these methods, however, is their computational complexity, especially when dealing with high-resolutionimages in which relatively long range interactions may play a role. We propose a new method based on Kriging thatmakes it possible to include such long range interactions, while keeping the computations manageable when dealingwith large medical images.",

author = "Melanie Ganz and Marco Loog and Sami Brandt and Mads Nielsen",

year = "2009",

doi = "10.1109/CVPR.2009.5204055",

language = "English",

isbn = "978-1-4244-3994-2",

pages = "87--93",

booktitle = "IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 2009. CVPR Workshops 2009.",

note = "IEEE Computer Society Workshop on Mathematical Methods in Biomedical Image Analysis (MMBIA 2009) ; Conference date: 20-06-2009 Through 25-06-2009",

}

TY - GEN

T1 - Dense Iterative Contextual Pixel Classification using Kriging

AU - Ganz, Melanie

AU - Loog, Marco

AU - Brandt, Sami

AU - Nielsen, Mads

N1 - Conference code: 10

PY - 2009

Y1 - 2009

N2 - In medical applications, segmentation has become an ever more important task. One of the competitive schemes toperform such segmentation is by means of pixel classification. Simple pixel-based classification schemes can be improved by incorporating contextual label information. Various methods have been proposed to this end, e.g., iterativecontextual pixel classification, iterated conditional modes, and other approaches related to Markov random fields. Aproblem of these methods, however, is their computational complexity, especially when dealing with high-resolutionimages in which relatively long range interactions may play a role. We propose a new method based on Kriging thatmakes it possible to include such long range interactions, while keeping the computations manageable when dealingwith large medical images.

AB - In medical applications, segmentation has become an ever more important task. One of the competitive schemes toperform such segmentation is by means of pixel classification. Simple pixel-based classification schemes can be improved by incorporating contextual label information. Various methods have been proposed to this end, e.g., iterativecontextual pixel classification, iterated conditional modes, and other approaches related to Markov random fields. Aproblem of these methods, however, is their computational complexity, especially when dealing with high-resolutionimages in which relatively long range interactions may play a role. We propose a new method based on Kriging thatmakes it possible to include such long range interactions, while keeping the computations manageable when dealingwith large medical images.

U2 - 10.1109/CVPR.2009.5204055

DO - 10.1109/CVPR.2009.5204055

M3 - Article in proceedings

SN - 978-1-4244-3994-2

SP - 87

EP - 93

BT - IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 2009. CVPR Workshops 2009.

T2 - IEEE Computer Society Workshop on Mathematical Methods in Biomedical Image Analysis (MMBIA 2009)

Y2 - 20 June 2009 through 25 June 2009

ER -

Dense Iterative Contextual Pixel Classification using Kriging

Abstract

Conference

Access to Document

Fingerprint

Cite this