emd.rar_EMD matlab_Matlab BE_emd 分解_分解

% Env_Rbf.m % BEMD % Interpolation by radial basis functions from extrema or crest lines % % J.C. Nunes, Y. Bouaoune, E. Del閏helle, N. Oumar, and Ph. Bunel. % "Image analysis by bidimensional empirical mode decomposition". % Image and Vision Computing Journal (IVC), (to appear), 2003. function [Ienvmin,Ienvmax,nbr_extrema]=Env_Rbf(I,ext_lpe) global name % si ext_lpe==1 interpolation par les extrema sinon par les lignes cr鑤es [t1 t2]=size(I); CONN=8; h=1; I2=double(I); %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Calcul de l'image des minima if ext_lpe==1 IMin=imregionalmin(I2); IMin=double(IMin); Imin=IMin.*I2; else I2_inv=double(imcomplement(double(I2)))*255; IMin=watershed(I2_inv,CONN); IMin=IMin==0; IMin=double(IMin); Imin=IMin.*I2; clear I2_inv end %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Calcul de l'image maxima if ext_lpe==1 IMax=imregionalmax(I2); IMax=double(IMax); Imax=IMax.*I2; else IMax=watershed(I2,CONN); IMax=IMax==0; IMax=double(IMax); Imax=IMax.*I2; end %%%%%%%%%%%%%%%%%%%%%% Les minima %%%%%%%%%%%%%%%%%%%%%%% % Calcul enveloppe min [x,y,z] = find(Imin); % on calcule l'env inf � partir des valeurs num閞iques min r間ionaux datamin.Location(1,:)=x(:); datamin.Location(2,:)=y(:); datamin.Value=z'; nbr_extrema=length(y); if length(y)<4 Ienvmin=I; Ienvmax=I; disp(['Pourquoi nbre de min si petit, nombre de min=' num2str(length(y))]) return else clear x y z rbfmin=fastrbf_fit(datamin,0.00000001); clear datamin gmin=fastrbf_grideval(rbfmin,'spacing',1,'size',[t1,t2],'accuracy',0.001); Ienvmin=gmin.Value; % enveloppe inf clear gmin end %%%%%%%%%%%%%%%%%%%%%% Les minima %%%%%%%%%%%%%%%%%%%%%%% % Calcul enveloppe max [x,y,z] = find(Imax); % on calcule l'env max � partir des valeurs num閞iques max r間ionaux datamax.Location(1,:)=x(:); datamax.Location(2,:)=y(:); datamax.Value=z'; nbr_extrema=nbr_extrema+length(y); if length(y)<4 Ienvmin=I; Ienvmax=I; disp(['Pourquoi nbre de max si petit, nombre de max=' num2str(length(y))]) return else clear x y z rbfmax=fastrbf_fit(datamax,0.00000001); clear datamax gmax=fastrbf_grideval(rbfmax,'spacing',1,'size',[t1,t2],'accuracy',0.001); Ienvmax=gmax.Value; clear gmax end clear gmax I2 rbfmax rbfmin IMax IMin return