LG.zip_图像多尺度_多尺度_多尺度 matlab_多尺度分析_多尺度图像

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Expand an image as per the Gaussian Pyramid. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% function IResult = GPExpand(I) Wt = [0.0500 0.2500 0.4000 0.2500 0.0500]; dim = size(I); newdim = dim*2; IResult = zeros(newdim,class(I)); % Initialize the array in the beginning .. m = [-2:2];n=m; switch length(dim) case 1 %% Pad the boundaries. I = [ I(1) ; I ; I(dim(1))]; for i = 0 : newdim(1) - 1 pixeli = (i - m)/2 + 2; idxi = find(floor(pixeli)==pixeli); A = I(pixeli(idxi)) .* Wt(m(idxi)+3); IResult(i + 1)= 2 * sum(A(:)); end case 2 %% Pad the boundaries. I = [ I(1,:) ; I ; I(dim(1),:) ]; % Pad the top and bottom rows. I = [ I(:,1) I I(:,dim(2)) ]; % Pad the left and right columns. Wt2 = Wt'*Wt; for i = 0 : newdim(1) - 1 for j = 0 : newdim(2) - 1 pixeli = (i - m)/2 + 2; idxi = find(floor(pixeli)==pixeli); pixelj = (j - m)/2 + 2; idxj = find(floor(pixelj)==pixelj); A = I(pixeli(idxi),pixelj(idxj)) .* Wt2(m(idxi)+3,m(idxj)+3); IResult(i + 1, j + 1)= 4 * sum(A(:)); end end case 3 Wt3 = ones(5,5,5); for i = 1:5 Wt3(i,:,:) = Wt3(i,:,:) * Wt(i); Wt3(:,i,:) = Wt3(:,i,:) * Wt(i); Wt3(:,:,i) = Wt3(:,:,i) * Wt(i); end %% Pad the boundaries I2 = zeros(dim+2,class(I)); I2(2:1+dim(1),2:1+dim(2),2:1+dim(3)) = I; I2(1,:,:)=I2(2,:,:);I2(end,:,:)=I2(end-1,:,:); I2(:,1,:)=I2(:,2,:);I2(:,end,:)=I2(:,end-1,:); I2(:,:,1)=I2(:,:,2);I2(:,:,end)=I2(:,:,end-1); I=I2; clear I2; for i = 0 : newdim(1) - 1 for j = 0 : newdim(2) - 1 for k = 0 : newdim(3) - 1 pixeli = (i - m)/2 + 2; idxi = find(floor(pixeli)==pixeli); pixelj = (j - m)/2 + 2; idxj = find(floor(pixelj)==pixelj); pixelk = (k - m)/2 + 2; idxk = find(floor(pixelk)==pixelk); A = I(pixeli(idxi),pixelj(idxj),pixelk(idxk)) .* Wt3(m(idxi)+3,m(idxj)+3,m(idxk)+3); IResult(i + 1, j + 1,k+1)= 8 * sum(A(:)); end end end end