Adaptive classification of textured images using linear prediction and neural networks

Sukissian, L; Kollias, S; Boutalis, Y

dc.contributor.author	Sukissian, L	en
dc.contributor.author	Kollias, S	en
dc.contributor.author	Boutalis, Y	en
dc.date.accessioned	2014-03-01T01:09:41Z
dc.date.available	2014-03-01T01:09:41Z
dc.date.issued	1994	en
dc.identifier.issn	0165-1684	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/11145
dc.subject	Classification	en
dc.subject	Itakura distance	en
dc.subject	Linear prediction	en
dc.subject	Neural networks	en
dc.subject	Segmentation	en
dc.subject	Texture	en
dc.subject.classification	Engineering, Electrical & Electronic	en
dc.subject.other	Adaptive systems	en
dc.subject.other	Algorithms	en
dc.subject.other	Computational complexity	en
dc.subject.other	Computer architecture	en
dc.subject.other	Distance measurement	en
dc.subject.other	Mathematical models	en
dc.subject.other	Neural networks	en
dc.subject.other	Parameter estimation	en
dc.subject.other	Recursive functions	en
dc.subject.other	Regression analysis	en
dc.subject.other	Spurious signal noise	en
dc.subject.other	Textures	en
dc.subject.other	Adaptive classification	en
dc.subject.other	Autoregressive texture model	en
dc.subject.other	Fast multichannel algorithm	en
dc.subject.other	Image classification	en
dc.subject.other	Image segmentation	en
dc.subject.other	Linear prediction	en
dc.subject.other	Statistical distance	en
dc.subject.other	Textured image	en
dc.subject.other	Image processing	en
dc.title	Adaptive classification of textured images using linear prediction and neural networks	en
heal.type	journalArticle	en
heal.identifier.primary	10.1016/0165-1684(94)90209-7	en
heal.identifier.secondary	http://dx.doi.org/10.1016/0165-1684(94)90209-7	en
heal.language	English	en
heal.publicationDate	1994	en
heal.abstract	A new technique for classifying and segmenting textured images is presented in this paper. This technique uses a fast multichannel algorithm for recursive estimation of autoregressive (AR) texture model parameters, along with a powerful statistical distance measure for adaptive selection of a small representative set of AR texture models. Specific properties of the estimation part of the method are exploited, which severly reduce the computational complexity of the distance measure, while robustness of classification with respect to additive noise is achieved using a higher-order version of the estimation part of the algorithm. Based on the small size of the selected set of AR texture models, appropriately designed and recursively constructed neural network architectures are included in the proposed scheme for adaptive classification and segmentation of textured images. © 1994.	en
heal.publisher	ELSEVIER SCIENCE BV	en
heal.journalName	Signal Processing	en
dc.identifier.doi	10.1016/0165-1684(94)90209-7	en
dc.identifier.isi	ISI:A1994NE15200007	en
dc.identifier.volume	36	en
dc.identifier.issue	2	en
dc.identifier.spage	209	en
dc.identifier.epage	232	en