Task-Based Evaulation of Image Filtering within a Class of Geometry-Driven-Diffusion Algorithms

Ivan Bajla; Igor Holländer; Viktor Witkovsky

Task-Based Evaulation of Image Filtering within a Class of Geometry-Driven-Diffusion Algorithms

Ivan Bajla
, Igor Holländer
, Viktor Witkovsky

Slovak Academy of Sciences

Research output: Chapter in Book or Conference Proceedings › Book chapter

Abstract

A novel task-based algorithm performance evaluation technique is propose for the evaluation of geometry-driven diffusion (GDD) methods used for increasing the signal-to-noise ratio in MR tomograms. It is based on a probabilistic model of stepwise constant image corrupted the distribution parameters of the ran-dom variable derived from intensity gradient are used for characterization of staircase image artifacts in diffused images. The proposed evaluation tech-nique incorporates a "gold standard" of the GDD algorithms, defined as a dif-fusion process governed by ideal values of conductance

Original language	English
Title of host publication	erschienen in: Empirical Evaluation Methods in Computer Vision. Series in machine perception and artificial intelligence. Ed. by: Christensen Henrik I.
Pages	89-115
Number of pages	27
Publication status	Published - 2002

Research Field

Not defined

Cite this

@inbook{4364534c5ce541da874b233085875600,

title = "Task-Based Evaulation of Image Filtering within a Class of Geometry-Driven-Diffusion Algorithms",

abstract = "A novel task-based algorithm performance evaluation technique is propose for the evaluation of geometry-driven diffusion (GDD) methods used for increasing the signal-to-noise ratio in MR tomograms. It is based on a probabilistic model of stepwise constant image corrupted the distribution parameters of the ran-dom variable derived from intensity gradient are used for characterization of staircase image artifacts in diffused images. The proposed evaluation tech-nique incorporates a {"}gold standard{"} of the GDD algorithms, defined as a dif-fusion process governed by ideal values of conductance",

author = "Ivan Bajla and Igor Holl{\"a}nder and Viktor Witkovsky",

year = "2002",

language = "English",

isbn = "981-02-4953-5",

pages = "89--115",

booktitle = "erschienen in: Empirical Evaluation Methods in Computer Vision. Series in machine perception and artificial intelligence. Ed. by: Christensen Henrik I.",

}

Task-Based Evaulation of Image Filtering within a Class of Geometry-Driven-Diffusion Algorithms. / Bajla, Ivan; Holländer, Igor; Witkovsky, Viktor.
erschienen in: Empirical Evaluation Methods in Computer Vision. Series in machine perception and artificial intelligence. Ed. by: Christensen Henrik I.. 2002. p. 89-115.

Research output: Chapter in Book or Conference Proceedings › Book chapter

TY - CHAP

T1 - Task-Based Evaulation of Image Filtering within a Class of Geometry-Driven-Diffusion Algorithms

AU - Bajla, Ivan

AU - Holländer, Igor

AU - Witkovsky, Viktor

PY - 2002

Y1 - 2002

N2 - A novel task-based algorithm performance evaluation technique is propose for the evaluation of geometry-driven diffusion (GDD) methods used for increasing the signal-to-noise ratio in MR tomograms. It is based on a probabilistic model of stepwise constant image corrupted the distribution parameters of the ran-dom variable derived from intensity gradient are used for characterization of staircase image artifacts in diffused images. The proposed evaluation tech-nique incorporates a "gold standard" of the GDD algorithms, defined as a dif-fusion process governed by ideal values of conductance

AB - A novel task-based algorithm performance evaluation technique is propose for the evaluation of geometry-driven diffusion (GDD) methods used for increasing the signal-to-noise ratio in MR tomograms. It is based on a probabilistic model of stepwise constant image corrupted the distribution parameters of the ran-dom variable derived from intensity gradient are used for characterization of staircase image artifacts in diffused images. The proposed evaluation tech-nique incorporates a "gold standard" of the GDD algorithms, defined as a dif-fusion process governed by ideal values of conductance

M3 - Book chapter

SN - 981-02-4953-5

SP - 89

EP - 115

BT - erschienen in: Empirical Evaluation Methods in Computer Vision. Series in machine perception and artificial intelligence. Ed. by: Christensen Henrik I.

ER -

Task-Based Evaulation of Image Filtering within a Class of Geometry-Driven-Diffusion Algorithms

Abstract

Research Field

Fingerprint

Cite this