Decompose image into subbands, denoise using BLS-GSM method, and recompose again.

　　　　fh = decomp_reconst(im,Nsc,filter,block,noise,parent,covariance,optim,sig)

im: image

Nsc: number of scales

filter: type of filter used (see namedFilters)

block: 2x1 vector indicating the dimensions (rows and columns) of the spatial neighborhood

noise: signal with the same autocorrelation as the noise

parent: include (1) or not (0) a coefficient from the immediately coarser scale in the neighborhood

covariance:　　 are we considering covariance or just variance?

optim:　　　　 for choosing between BLS-GSM (optim = 1) and MAP-GSM (optim = 0)

sig: standard deviation (scalar for uniform noise or matrix for spatially varying noise)

Version using a critically sampled pyramid (orthogonal wavelet), as implemented in MatlabPyrTools (Eero).



JPM, Univ. de Granada, 3/03-

　Decompose image into subbands, denoise using BLS-GSM method, and recompose again.

　　　　fh = decomp_reconst(im,Nsc,filter,block,noise,parent,covariance,optim,sig)

　　　　im:　　　　 image

　　　　Nsc:　　　　number of scales

　　　　filter:　　 type of filter used (see namedFilters)

　　　　block:　　　2x1 vector indicating the dimensions (rows and columns) of the spatial neighborhood

　　　　noise:　　　signal with the same autocorrelation as the noise

　　　　parent:　　 include (1) or not (0) a coefficient from the immediately coarser scale in the neighborhood

　　　　covariance:　　 are we considering covariance or just variance?

　　　　optim:　　　　 for choosing between BLS-GSM (optim = 1) and MAP-GSM (optim = 0)

　　　　sig:　　　　standard deviation (scalar for uniform noise or matrix for spatially varying noise)

　Version using a critically sampled pyramid (orthogonal wavelet), as implemented in MatlabPyrTools (Eero).



　JPM, Univ. de Granada, 3/03