Soft thresholding by noise invalidation

A new thresholding technique for data denoising is proposed. Using statistical properties of additive noise, this method provides an adaptive data dependent soft threshold to remove the effects of noise. The observed data can be denoised in any orthogonal basis. The simulations demonstrate the advan...

Full description

Saved in:
Bibliographic Details
Main Authors: Beheshti, S., Nikvand, N., Fernando, X.N.
Format: Conference Proceeding
Language:English
Subjects:
Additive noise
Gaussian noise
Noise reduction
Random processes
Statistics
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A new thresholding technique for data denoising is proposed. Using statistical properties of additive noise, this method provides an adaptive data dependent soft threshold to remove the effects of noise. The observed data can be denoised in any orthogonal basis. The simulations demonstrate the advantages of the new approach for data denoising with wavelet transformation.
DOI:10.1109/BSC.2008.4563246