Spectral-morphological analysis of acoustical images of biological tissues and composite structures: II. Test classification of acoustomicroscopic images

Autor:	Valentin A. Burov, Olga D. Rumyantseva, E. L. Kim
Rok vydání:	2005
Předmět:	Acoustics and Ultrasonics Computer science business.industry Bayesian probability Composite number Structure (category theory) Spectral density Pattern recognition Structure type Morphological analysis Structures Artificial intelligence business Statistical hypothesis testing
Zdroj:	Acoustical Physics. 51:255-264
ISSN:	1562-6865 1063-7710
DOI:	10.1134/1.1922538
Popis:	Results of numerical classification of different structure types in acoustomicroscopic images of real biological tissues and composite structures are reported. The classification is based on the Bayesian statistical hypothesis testing method. General realizations of this method in the form of the spatial-spectrum and spectral-morphological approaches are considered. The spatial-spectrum approach cannot discriminate between morphologically different structures with statistically identical power spectra. This disadvantage is eliminated in the spectral-morphological approach, which uses information on the spectral properties of the structure type being classified, as well as on the particular form of its characteristic features. The general approaches and their particular realizations (Bartlett’s and Pisarenko’s methods) are compared as applied to practical problems.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::6deac2eab0f4751a8cd41929a4597dec https://doi.org/10.1134/1.1922538 Zobrazit plný text záznamu Plný text ve formátu PDF