Blind deconvolution of natural images using segmentation based CMA

Pradeepa D. Samarasinghe; Rodney A. Kennedy

doi:10.1109/ICSPCS.2010.5709712

Blind deconvolution of natural images using segmentation based CMA

Pradeepa D. Samarasinghe, Rodney A. Kennedy

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

3 Citations (Scopus)

Abstract

In this paper, we analyze the applicability of Constant Modulus Algorithm (CMA), one of the most widely used and tested blind equalization technique to blind image deconvolution. With a detailed mathematical analysis, we show that the strong correlation between the neighboring spatial locations found in natural images becomes a major constraint on the convergence of CMA. In order to overcome this constraint, we introduce a novel image pixel correlation model in relation with natural image statistics. Based on this model, a segmented blind image deconvolution through CMA is proposed. The robustness of the proposed algorithm with natural images is discussed in terms of efficiency and effectiveness.

Original language	English
Title of host publication	4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings
DOIs	https://doi.org/10.1109/ICSPCS.2010.5709712
Publication status	Published - 2010
Event	4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Gold Coast, QLD, Australia Duration: 13 Dec 2010 → 15 Dec 2010

Publication series

Name	4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings

Conference

Conference	4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010
Country/Territory	Australia
City	Gold Coast, QLD
Period	13/12/10 → 15/12/10

Access to Document

10.1109/ICSPCS.2010.5709712

Cite this

Samarasinghe, P. D., & Kennedy, R. A. (2010). Blind deconvolution of natural images using segmentation based CMA. In 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings Article 5709712 (4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings). https://doi.org/10.1109/ICSPCS.2010.5709712

@inproceedings{8e6edfaf728c495ba868fea821d3ae58,

title = "Blind deconvolution of natural images using segmentation based CMA",

abstract = "In this paper, we analyze the applicability of Constant Modulus Algorithm (CMA), one of the most widely used and tested blind equalization technique to blind image deconvolution. With a detailed mathematical analysis, we show that the strong correlation between the neighboring spatial locations found in natural images becomes a major constraint on the convergence of CMA. In order to overcome this constraint, we introduce a novel image pixel correlation model in relation with natural image statistics. Based on this model, a segmented blind image deconvolution through CMA is proposed. The robustness of the proposed algorithm with natural images is discussed in terms of efficiency and effectiveness.",

keywords = "Blind image deconvolution, Constant modulus algorithm, Equalization, Image correlation, Kurtosis, Meso-Kurtic, Natural image statistics, Stationary points, Whitening",

author = "Samarasinghe, {Pradeepa D.} and Kennedy, {Rodney A.}",

year = "2010",

doi = "10.1109/ICSPCS.2010.5709712",

language = "English",

isbn = "9781424479078",

series = "4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings",

booktitle = "4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings",

note = "4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 ; Conference date: 13-12-2010 Through 15-12-2010",

}

Samarasinghe, PD & Kennedy, RA 2010, Blind deconvolution of natural images using segmentation based CMA. in 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings., 5709712, 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings, 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010, Gold Coast, QLD, Australia, 13/12/10. https://doi.org/10.1109/ICSPCS.2010.5709712

Blind deconvolution of natural images using segmentation based CMA. / Samarasinghe, Pradeepa D.; Kennedy, Rodney A.
4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings. 2010. 5709712 (4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings).

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

TY - GEN

T1 - Blind deconvolution of natural images using segmentation based CMA

AU - Samarasinghe, Pradeepa D.

AU - Kennedy, Rodney A.

PY - 2010

Y1 - 2010

N2 - In this paper, we analyze the applicability of Constant Modulus Algorithm (CMA), one of the most widely used and tested blind equalization technique to blind image deconvolution. With a detailed mathematical analysis, we show that the strong correlation between the neighboring spatial locations found in natural images becomes a major constraint on the convergence of CMA. In order to overcome this constraint, we introduce a novel image pixel correlation model in relation with natural image statistics. Based on this model, a segmented blind image deconvolution through CMA is proposed. The robustness of the proposed algorithm with natural images is discussed in terms of efficiency and effectiveness.

AB - In this paper, we analyze the applicability of Constant Modulus Algorithm (CMA), one of the most widely used and tested blind equalization technique to blind image deconvolution. With a detailed mathematical analysis, we show that the strong correlation between the neighboring spatial locations found in natural images becomes a major constraint on the convergence of CMA. In order to overcome this constraint, we introduce a novel image pixel correlation model in relation with natural image statistics. Based on this model, a segmented blind image deconvolution through CMA is proposed. The robustness of the proposed algorithm with natural images is discussed in terms of efficiency and effectiveness.

KW - Blind image deconvolution

KW - Constant modulus algorithm

KW - Equalization

KW - Image correlation

KW - Kurtosis

KW - Meso-Kurtic

KW - Natural image statistics

KW - Stationary points

KW - Whitening

UR - http://www.scopus.com/inward/record.url?scp=79952506607&partnerID=8YFLogxK

U2 - 10.1109/ICSPCS.2010.5709712

DO - 10.1109/ICSPCS.2010.5709712

M3 - Conference contribution

SN - 9781424479078

T3 - 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings

BT - 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010 - Proceedings

T2 - 4th International Conference on Signal Processing and Communication Systems, ICSPCS'2010

Y2 - 13 December 2010 through 15 December 2010

ER -

Blind deconvolution of natural images using segmentation based CMA

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this