Title :
Texture synthesis via a noncausal nonparametric multiscale Markov random field
Author :
Paget, Rupert ; Longstaff, I. Dennis
Author_Institution :
Dept. of Electr. & Comput. Eng., Queensland Univ., Brisbane, Qld., Australia
fDate :
6/1/1998 12:00:00 AM
Abstract :
Our noncausal, nonparametric, multiscale, Markov random field (MRF) model is capable of synthesizing and capturing the characteristics of a wide variety of textures, from the highly structured to the stochastic. We use a multiscale synthesis algorithm incorporating local annealing to obtain larger realizations of texture visually indistinguishable from the training texture
Keywords :
Markov processes; higher order statistics; image classification; image texture; nonparametric statistics; Markov random field; high-order statistical characteristics; image texture synthesis; local annealing; multiscale synthesis algorithm; noncausal nonparametric multiscale model; Annealing; Autoregressive processes; Fractals; Image segmentation; Image texture analysis; Markov random fields; Multidimensional systems; Sensor phenomena and characterization; Stochastic processes; Synthetic aperture radar;
Journal_Title :
Image Processing, IEEE Transactions on
