Modified noncausal smoothing filter and low rank matrix approximation for noise reduction

Teeradache Viangteeravat, D. M. Wilkes

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Removing noise in real time has become a high priority for analyzing data corrupted by additive noise. It is a major problem in various applications such as speech, image processing and real time multimedia services. Although considerable interest has arisen in recent years regarding wavelets as a new transform technique for many applications, the linear adaptive decomposition transform (LDT) has yielded results superior to the discrete wavelet transform (DWT) not only in terms of using a lower number of decomposition levels but also achieving a smaller percentage normalized approximation error in the reconstructed signal. In this paper, a novel noise reduction method, based on a modified noncausal smoothing filter and low rank approximation based upon the sum of minimum magnitude error criterion (i.e., l1 norm) is introduced that distinguishes itself from these other methods. The performance of the proposed approach was evaluated based on one dimensional data sets as well as speech samples. It is demonstrated that the approach yields very promising results on the test signals of the Donoho and Johnstone as well as to speech signals.

Original languageEnglish (US)
Title of host publicationSignal Processing, Sensor Fusion, and Target Recognition XIX
DOIs
StatePublished - Jun 18 2010
EventSignal Processing, Sensor Fusion, and Target Recognition XIX - Orlando, FL, United States
Duration: Apr 5 2010Apr 7 2010

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume7697
ISSN (Print)0277-786X

Other

OtherSignal Processing, Sensor Fusion, and Target Recognition XIX
CountryUnited States
CityOrlando, FL
Period4/5/104/7/10

Fingerprint

Matrix Approximation
Low-rank Approximation
Low-rank Matrices
Noise Reduction
Noise abatement
noise reduction
smoothing
Smoothing
Transform
Filter
filters
Decompose
L1-norm
Speech Signal
Additive Noise
Approximation Error
Reduction Method
approximation
Wavelet Transform
Decomposition

All Science Journal Classification (ASJC) codes

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

Viangteeravat, T., & Wilkes, D. M. (2010). Modified noncausal smoothing filter and low rank matrix approximation for noise reduction. In Signal Processing, Sensor Fusion, and Target Recognition XIX [76971K] (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 7697). https://doi.org/10.1117/12.858034

Modified noncausal smoothing filter and low rank matrix approximation for noise reduction. / Viangteeravat, Teeradache; Wilkes, D. M.

Signal Processing, Sensor Fusion, and Target Recognition XIX. 2010. 76971K (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 7697).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Viangteeravat, T & Wilkes, DM 2010, Modified noncausal smoothing filter and low rank matrix approximation for noise reduction. in Signal Processing, Sensor Fusion, and Target Recognition XIX., 76971K, Proceedings of SPIE - The International Society for Optical Engineering, vol. 7697, Signal Processing, Sensor Fusion, and Target Recognition XIX, Orlando, FL, United States, 4/5/10. https://doi.org/10.1117/12.858034
Viangteeravat T, Wilkes DM. Modified noncausal smoothing filter and low rank matrix approximation for noise reduction. In Signal Processing, Sensor Fusion, and Target Recognition XIX. 2010. 76971K. (Proceedings of SPIE - The International Society for Optical Engineering). https://doi.org/10.1117/12.858034
Viangteeravat, Teeradache ; Wilkes, D. M. / Modified noncausal smoothing filter and low rank matrix approximation for noise reduction. Signal Processing, Sensor Fusion, and Target Recognition XIX. 2010. (Proceedings of SPIE - The International Society for Optical Engineering).
@inproceedings{8d329a7dd7c841db9fdadaea7bdd730c,
title = "Modified noncausal smoothing filter and low rank matrix approximation for noise reduction",
abstract = "Removing noise in real time has become a high priority for analyzing data corrupted by additive noise. It is a major problem in various applications such as speech, image processing and real time multimedia services. Although considerable interest has arisen in recent years regarding wavelets as a new transform technique for many applications, the linear adaptive decomposition transform (LDT) has yielded results superior to the discrete wavelet transform (DWT) not only in terms of using a lower number of decomposition levels but also achieving a smaller percentage normalized approximation error in the reconstructed signal. In this paper, a novel noise reduction method, based on a modified noncausal smoothing filter and low rank approximation based upon the sum of minimum magnitude error criterion (i.e., l1 norm) is introduced that distinguishes itself from these other methods. The performance of the proposed approach was evaluated based on one dimensional data sets as well as speech samples. It is demonstrated that the approach yields very promising results on the test signals of the Donoho and Johnstone as well as to speech signals.",
author = "Teeradache Viangteeravat and Wilkes, {D. M.}",
year = "2010",
month = "6",
day = "18",
doi = "10.1117/12.858034",
language = "English (US)",
isbn = "9780819481610",
series = "Proceedings of SPIE - The International Society for Optical Engineering",
booktitle = "Signal Processing, Sensor Fusion, and Target Recognition XIX",

}

TY - GEN

T1 - Modified noncausal smoothing filter and low rank matrix approximation for noise reduction

AU - Viangteeravat, Teeradache

AU - Wilkes, D. M.

PY - 2010/6/18

Y1 - 2010/6/18

N2 - Removing noise in real time has become a high priority for analyzing data corrupted by additive noise. It is a major problem in various applications such as speech, image processing and real time multimedia services. Although considerable interest has arisen in recent years regarding wavelets as a new transform technique for many applications, the linear adaptive decomposition transform (LDT) has yielded results superior to the discrete wavelet transform (DWT) not only in terms of using a lower number of decomposition levels but also achieving a smaller percentage normalized approximation error in the reconstructed signal. In this paper, a novel noise reduction method, based on a modified noncausal smoothing filter and low rank approximation based upon the sum of minimum magnitude error criterion (i.e., l1 norm) is introduced that distinguishes itself from these other methods. The performance of the proposed approach was evaluated based on one dimensional data sets as well as speech samples. It is demonstrated that the approach yields very promising results on the test signals of the Donoho and Johnstone as well as to speech signals.

AB - Removing noise in real time has become a high priority for analyzing data corrupted by additive noise. It is a major problem in various applications such as speech, image processing and real time multimedia services. Although considerable interest has arisen in recent years regarding wavelets as a new transform technique for many applications, the linear adaptive decomposition transform (LDT) has yielded results superior to the discrete wavelet transform (DWT) not only in terms of using a lower number of decomposition levels but also achieving a smaller percentage normalized approximation error in the reconstructed signal. In this paper, a novel noise reduction method, based on a modified noncausal smoothing filter and low rank approximation based upon the sum of minimum magnitude error criterion (i.e., l1 norm) is introduced that distinguishes itself from these other methods. The performance of the proposed approach was evaluated based on one dimensional data sets as well as speech samples. It is demonstrated that the approach yields very promising results on the test signals of the Donoho and Johnstone as well as to speech signals.

UR - http://www.scopus.com/inward/record.url?scp=77953533487&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77953533487&partnerID=8YFLogxK

U2 - 10.1117/12.858034

DO - 10.1117/12.858034

M3 - Conference contribution

SN - 9780819481610

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - Signal Processing, Sensor Fusion, and Target Recognition XIX

ER -