Optimum Filtering as Multicriteria Decision Making Task

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F60461373%3A22340%2F02%3A00007059" target="_blank" >RIV/60461373:22340/02:00007059 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
Optimum Filtering as Multicriteria Decision Making Task
Original language description
Filtering quality can be estimated on various data sets by various measures. Five basic quality measures were used: mean square of error (MSE), mean of absolute error (MAE), maximum of absolute error (MAX), median of absolute error (MED) and signal to noise ratio (SNR). Then the quality of given filter is not a scalar but it is a vector of various measures on various data sets. The testing results of various filters can be collected to quality matrix. There is a possibility to find optimum filter usingmulticriteria decision-making technique. Finding ideal and basic alternatives is the first step. The quality matrix normalization is the second step. The last step is based on no a priori knowledge of quality measure weights and we use AIA technique forthe best filter selection. The resulting optimum filter guarantees compromise quality of various data set processing. The general principle is demonstrated on 2D MRI image de-noising.
Czech name
—
Czech description
—

Publication year
2002
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection
Proc. of the 5th Int. Scientific - Technical Conference on Process Control '02
ISBN
80-7194-452-1
ISSN
—
e-ISSN
—
Number of pages
8
Pages from-to
—
Publisher name
University of Pardubice
Place of publication
Pardubice
Event location
Kouty nad Desnou
Event date
Jun 9, 2002
Type of event by nationality
EUR - Evropská akce
UT code for WoS article
—

Similar results(10)