DocumentCode :
1281139
Title :
Multiscale signal enhancement: beyond the normality and independence assumption
Author :
He, Yun ; Krim, Hamid
Author_Institution :
Analog/Mixed Signal IC Design Group, Tality Corp., Cary, NC, USA
Volume :
11
Issue :
4
fYear :
2002
fDate :
4/1/2002 12:00:00 AM
Firstpage :
423
Lastpage :
433
Abstract :
Current approaches to denoising or signal enhancement in a wavelet-based framework have generally relied on the assumption of normally distributed perturbations. In practice, this assumption is often violated and sometimes prior information of the probability distribution of a noise process is not even available. To relax this assumption, we propose a novel nonlinear filtering technique in this paper. The key idea is to project a noisy signal onto a wavelet domain and to suppress wavelet coefficients by a mask derived from curvature extrema in its scale space representation. For a piecewise smooth signal, it can be shown that filtering by this curvature mask is equivalent to preserving the signal pointwise Holder exponents at the singular points and lifting its smoothness elsewhere
Keywords :
image enhancement; noise; nonlinear filters; probability; smoothing methods; wavelet transforms; curvature extrema; curvature mask filtering; human visual system; independence assumption; multiscale analysis; multiscale signal enhancement; noise process; noisy image; nonlinear filtering; normality assumption; normally distributed perturbations; piecewise smooth signal; probability distribution; scale space representation; signal pointwise Holder exponents; singular points; wavelet coefficients suppression; wavelet domain; wavelet transform; Filtering; Image analysis; Information analysis; Minimax techniques; Noise reduction; Probability distribution; Signal analysis; Signal processing; Wavelet analysis; Wavelet coefficients;
fLanguage :
English
Journal_Title :
Image Processing, IEEE Transactions on
Publisher :
ieee
ISSN :
1057-7149
Type :
jour
DOI :
10.1109/TIP.2002.999676
Filename :
999676
Link To Document :
بازگشت