Skryté Markovovy stromy wavelet koeficientů

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F60461373%3A22340%2F08%3A00020849" target="_blank" >RIV/60461373:22340/08:00020849 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Wavelet-Based Hidden Markov Trees for Noise Reduction

Popis výsledku v původním jazyce

In the field of signal processing, the Discrete Wavelet Transform (DWT) has proved very useful for recovering signals from additive Gaussian noise by the means of wavelet thresholding. During this procedure, wavelet coefficients with small magnitudes areset to zero, however, usually without taking into account their mutual dependencies. The Hidden Markov Models (HMM) are designed to capture such dependencies by modelling the statistical properties of the coefficients. In this paper, we process a testing intensity image with added Gaussian noise. To compute the hidden Markov models parameters, we employ the iterative expectation-maximization (EM) training algorithm. The outcome of the training process is used for estimation of the noise-free image which is reconstructed from the recalculated wavelet coefficients. The above technique is compared with the NormalShrink method of adaptive threshold computation and outperforms this technique in our experiments.

Název v anglickém jazyce

Wavelet-Based Hidden Markov Trees for Noise Reduction

Popis výsledku anglicky

In the field of signal processing, the Discrete Wavelet Transform (DWT) has proved very useful for recovering signals from additive Gaussian noise by the means of wavelet thresholding. During this procedure, wavelet coefficients with small magnitudes areset to zero, however, usually without taking into account their mutual dependencies. The Hidden Markov Models (HMM) are designed to capture such dependencies by modelling the statistical properties of the coefficients. In this paper, we process a testing intensity image with added Gaussian noise. To compute the hidden Markov models parameters, we employ the iterative expectation-maximization (EM) training algorithm. The outcome of the training process is used for estimation of the noise-free image which is reconstructed from the recalculated wavelet coefficients. The above technique is compared with the NormalShrink method of adaptive threshold computation and outperforms this technique in our experiments.

Druh

D - Stať ve sborníku

CEP obor

JD - Využití počítačů, robotika a její aplikace

OECD FORD obor

—

Projekt

—

Návaznosti

Z - Vyzkumny zamer (s odkazem do CEZ)

Rok uplatnění

2008

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

16th Annual Conf. Proc. Techn. Computing Prague 2008

ISBN

978-80-7080-692-0

ISSN

—

e-ISSN

—

Počet stran výsledku

10

Strana od-do

—

Název nakladatele

Humusoft

Místo vydání

Praha

Místo konání akce

Praha

Datum konání akce

11. 11. 2008

Typ akce podle státní příslušnosti

EUR - Evropská akce

Kód UT WoS článku

—