Detail publikace

Generalized Cross Validation with Bayes Approach for Image Denoising

ČERVINKA, T.

Originální název

Generalized Cross Validation with Bayes Approach for Image Denoising

Typ

článek ve sborníku ve WoS nebo Scopus

Jazyk

angličtina

Originální abstrakt

This paper describes a new method for suppression of noise in images based on wavelet transform. Classic de-noising methods based on the wavelet transform are based on binary decision. Wavelet coefficients below threshold are replaced by zero, and kept (hard-threshold) or shrinked (soft-threshold) with absolute value above the threshold. The proposed method relies on two criteria. The first criterion is based on estimation of optimal threshold without knowledge exact data using Generalized Cross Validation (GCV) technique. The second criterion employes Bayes approach to improve noise suppression in images.

Klíčová slova

image processing, noise suppression, Markov Random Fields, applied probability, Hőlder regularity, random number, Generalized Cross Validation

Autoři

ČERVINKA, T.

Rok RIV

2004

Vydáno

1. 1. 2004

Místo

Praha

Strany od

20

Strany do

25

Strany počet

6

URL

http://radio.feld.cvut.cz/conf/poster2004/

BibTex

@inproceedings{BUT13290,
  author="Tomáš {Červinka}",
  title="Generalized Cross Validation with Bayes Approach for Image Denoising",
  booktitle="8th International Student Conference on Electrical Engineering",
  year="2004",
  number="1",
  pages="6",
  address="Praha",
  url="http://radio.feld.cvut.cz/conf/poster2004/"
}