0410383 - UTIA-B 20000099 RIV DE eng C - Conference Paper (international conference)
Haindl, Michal - Havlíček, Vojtěch
A multiresolution causal colour texture model.
Berlin: Springer, 2000. Lecture Notes in Computer Science., 1876. ISBN 3-540-67946-4. In: Advances in Pattern Recognition. Proceedings. - (Ferri, F.; Inesta, J.; Amin, A.; Pudil, P.), s. 114-122
[Joint IAPR International Workshops SSPR 2000 and SPR 2000. Alicante (ES), 30.08.2000-01.09.2000]
R&D Projects: GA ČR GA102/00/0030; GA ČR GA106/00/1715
Grant - others:Hewlett Packard(US) ISE-86L8-13
Institutional research plan: AV0Z1075907
Subject RIV: BD - Theory of Information

An efficient recursive algorithm for realistic colour texture synthesis is proposed. The algorithm starts with spectral factorization of an input colour texture image using the Karhunen-Loeve decorrelation. Single orthogonal monospectral components are further decomposed into a multi-resolution grid and each resolution data are independently modeled by their dedicated simultaneous causal autoregressive random field model (CAR). We estimate an optimal contextual neighbourhood and model parameters.
Permanent Link: http://hdl.handle.net/11104/0130472