An FFT implementation of the generalized maximum likelihood algorithm for image smoothing

Author

Göksel, N. Sibel ; Namazi, Nader M.

Author_Institution

Dept. of Electr. Eng., Michigan Technol. Univ., Houghton, MI, USA

fYear

1990

fDate

12-14 Aug 1990

Firstpage

1139

Abstract

Practical implementation of the generalized maximum-likelihood algorithm on noise-corrupted images becomes prohibitive when the covariance of the noise-free image is unavailable. By partitioning the image into locally Markovian sub-blocks with separable correlation coefficients, a covariance model is found that enables fast Fourier transform (FFT) processing. Simulations using real images are presented to characterize the algorithm´s applicability and noise-reduction performance

Keywords

fast Fourier transforms; picture processing; probability; FFT implementation; covariance model; fast Fourier transform; generalized maximum likelihood algorithm; image smoothing; locally Markovian sub-blocks; noise-corrupted images; noise-reduction performance; partitioning; separable correlation coefficients; Anisotropic magnetoresistance; Computational modeling; Convergence; Convolution; Covariance matrix; Flowcharts; Kernel; Noise figure; Smoothing methods; Statistics;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 1990., Proceedings of the 33rd Midwest Symposium on

Conference_Location

Calgary, Alta.

Print_ISBN

0-7803-0081-5

Type

conf

DOI

10.1109/MWSCAS.1990.140927

Filename

140927

Link To Document

https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=2832418