Title :
Image classification using spectral and spatial information based on MRF models
Author :
Yamazaki, Tatsuya ; Gingras, Denis
Author_Institution :
Commun. Res. Lab., Kansai Adv. Res. Center, Kobe, Japan
fDate :
9/1/1995 12:00:00 AM
Abstract :
A new criterion for classifying multispectral remote sensing images or textured images by using spectral and spatial information is proposed. The images are modeled with a hierarchical Markov Random Field (MRF) model that consists of the observed intensity process and the hidden class label process. The class labels are estimated according to the maximum a posteriori (MAP) criterion, but some reasonable approximations are used to reduce the computational load. A stepwise classification algorithm is derived and is confirmed by simulation and experimental results
Keywords :
Markov processes; approximation theory; image classification; image texture; maximum likelihood estimation; random processes; remote sensing; spectral analysis; MAP estimation; MRF models; approximations; experimental results; hidden class label process; hierarchical Markov random field model; image classification; maximum a posteriori criterion; multispectral remote sensing images; observed intensity process; simulation results; spatial information; spectral information; stepwise classification algorithm; textured images; Classification algorithms; Computational modeling; Image classification; Image segmentation; Laboratories; Markov random fields; Optical sensors; Pixel; Remote sensing; Satellites;
Journal_Title :
Image Processing, IEEE Transactions on
