Title :
Bayesian classification of multivariate image after MAP reconstruction of noisy channels
Author :
Jhung, Yonhong ; Swain, Philip H.
Author_Institution :
Sch. of Electr. Eng., Purdue Univ., West Lafayette, IN, USA
Abstract :
Presents a supervised Bayesian classifier that makes use of both spectral signatures and spatial interactions after the preprocessing of clean noisy channels. The authors apply the Markov random field model at both preprocessing and classification stages. They perform the optimization using either coordinate descent or iterated conditional mode. The estimation of filter parameters is accomplished by referring to adjacent channels that have higher signal-to-noise ratio
Keywords :
Bayes methods; Markov processes; filtering and prediction theory; image recognition; optimisation; parameter estimation; remote sensing; Bayesian classification; MAP reconstruction; Markov random field model; adjacent channels; classification; coordinate descent; filter parameters; iterated conditional mode; maximum a posteriori estimate; multivariate image; noisy channels; optimization; preprocessing; signal-to-noise ratio; spatial interactions; spectral signatures; Bayesian methods; Cleaning; Filters; Image reconstruction; Markov random fields; Parameter estimation; Remote sensing; Signal to noise ratio; Stochastic processes; Working environment noise;
Conference_Titel :
System Theory, 1994., Proceedings of the 26th Southeastern Symposium on
Conference_Location :
Athens, OH
Print_ISBN :
0-8186-5320-5
DOI :
10.1109/SSST.1994.287840
