Multi-scale Image Decomposition using a Local Statistical Edge Model

Kin-Ming Wong

doi:10.1109/ICVR51878.2021.9483837

Multi-scale Image Decomposition using a Local Statistical Edge Model

School of Creative Media

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

11 Citations (Scopus)

Abstract

We present a progressive image decomposition method based on a novel non-linear filter named Sub-window Variance filter. Our method is specifically designed for image detail enhancement purpose; this application requires extraction of image details which are small in terms of both spatial and variation scales. We propose a local statistical edge model which develops its edge awareness using spatially defined image statistics. Our decomposition method is controlled by two intuitive parameters which allow the users to define what image details to suppress or enhance. By using the summed-area table acceleration method, our decomposition pipeline is highly parallel. The proposed filter is gradient preserving and this allows our enhancement results free from the gradient-reversal artefact. In our evaluations, we compare our method in various multi-scale image detail manipulation applications with other mainstream solutions.

Original language	English
Title of host publication	2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021)
Publisher	IEEE
Pages	10-18
Number of pages	9
ISBN (Electronic)	9781665423083, 9781665423090
ISBN (Print)	9781665446211
DOIs	https://doi.org/10.1109/ICVR51878.2021.9483837
Publication status	Published - 2021
Event	2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021) - Swissotel Foshan, Foshan, China Duration: 20 May 2021 → 22 May 2021 Conference number: 7 http://www.icvr.org/

Publication series

Name	International Conference on Virtual Rehabilitation, ICVR
Volume	2021-May
ISSN (Print)	2331-9542
ISSN (Electronic)	2331-9569

Conference

Conference	2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021)
Abbreviated title	ICVR 2021
Place	China
City	Foshan
Period	20/05/21 → 22/05/21
Internet address	http://www.icvr.org/

Bibliographical note

Information for this record is supplemented by the author(s) concerned.

Research Keywords

multi-scale image decomposition
image processing
edge-preserving filter

Access Output

10.1109/ICVR51878.2021.9483837

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@inproceedings{9c41296dd0064ab7a4fae2c8013ff4c3,

title = "Multi-scale Image Decomposition using a Local Statistical Edge Model",

abstract = "We present a progressive image decomposition method based on a novel non-linear filter named Sub-window Variance filter. Our method is specifically designed for image detail enhancement purpose; this application requires extraction of image details which are small in terms of both spatial and variation scales. We propose a local statistical edge model which develops its edge awareness using spatially defined image statistics. Our decomposition method is controlled by two intuitive parameters which allow the users to define what image details to suppress or enhance. By using the summed-area table acceleration method, our decomposition pipeline is highly parallel. The proposed filter is gradient preserving and this allows our enhancement results free from the gradient-reversal artefact. In our evaluations, we compare our method in various multi-scale image detail manipulation applications with other mainstream solutions.",

keywords = "multi-scale image decomposition, image processing, edge-preserving filter",

author = "Kin-Ming Wong",

note = "Information for this record is supplemented by the author(s) concerned.; 2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021), ICVR 2021 ; Conference date: 20-05-2021 Through 22-05-2021",

year = "2021",

doi = "10.1109/ICVR51878.2021.9483837",

language = "English",

isbn = "9781665446211",

series = "International Conference on Virtual Rehabilitation, ICVR",

publisher = "IEEE",

pages = "10--18",

booktitle = "2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021)",

address = "United States",

url = "http://www.icvr.org/",

}

Wong, K-M 2021, Multi-scale Image Decomposition using a Local Statistical Edge Model. in 2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021)., 9483837, International Conference on Virtual Rehabilitation, ICVR, vol. 2021-May, IEEE, pp. 10-18, 2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021), Foshan, China, 20/05/21. https://doi.org/10.1109/ICVR51878.2021.9483837

Multi-scale Image Decomposition using a Local Statistical Edge Model. / Wong, Kin-Ming.
2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021). IEEE, 2021. p. 10-18 9483837 (International Conference on Virtual Rehabilitation, ICVR; Vol. 2021-May).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - Multi-scale Image Decomposition using a Local Statistical Edge Model

AU - Wong, Kin-Ming

N1 - Conference code: 7

PY - 2021

Y1 - 2021

N2 - We present a progressive image decomposition method based on a novel non-linear filter named Sub-window Variance filter. Our method is specifically designed for image detail enhancement purpose; this application requires extraction of image details which are small in terms of both spatial and variation scales. We propose a local statistical edge model which develops its edge awareness using spatially defined image statistics. Our decomposition method is controlled by two intuitive parameters which allow the users to define what image details to suppress or enhance. By using the summed-area table acceleration method, our decomposition pipeline is highly parallel. The proposed filter is gradient preserving and this allows our enhancement results free from the gradient-reversal artefact. In our evaluations, we compare our method in various multi-scale image detail manipulation applications with other mainstream solutions.

AB - We present a progressive image decomposition method based on a novel non-linear filter named Sub-window Variance filter. Our method is specifically designed for image detail enhancement purpose; this application requires extraction of image details which are small in terms of both spatial and variation scales. We propose a local statistical edge model which develops its edge awareness using spatially defined image statistics. Our decomposition method is controlled by two intuitive parameters which allow the users to define what image details to suppress or enhance. By using the summed-area table acceleration method, our decomposition pipeline is highly parallel. The proposed filter is gradient preserving and this allows our enhancement results free from the gradient-reversal artefact. In our evaluations, we compare our method in various multi-scale image detail manipulation applications with other mainstream solutions.

KW - multi-scale image decomposition

KW - image processing

KW - edge-preserving filter

UR - http://www.scopus.com/inward/record.url?scp=85111422174&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85111422174&origin=recordpage

U2 - 10.1109/ICVR51878.2021.9483837

DO - 10.1109/ICVR51878.2021.9483837

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 9781665446211

T3 - International Conference on Virtual Rehabilitation, ICVR

SP - 10

EP - 18

BT - 2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021)

PB - IEEE

T2 - 2021 IEEE 7th International Conference on Virtual Reality (ICVR 2021)

Y2 - 20 May 2021 through 22 May 2021

ER -

Multi-scale Image Decomposition using a Local Statistical Edge Model

Abstract

Publication series

Conference

Bibliographical note

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this