武汉理工大学数字图像处理英文课件 Chapter4（4）.pptVIP

Colors in flowers.tif White component Color component Grayscale vs White Component Principal Components EIG Eigenvalues and Eigenvectors. Eigen-image 1 Eigen-image 2 Eigen-image 3 * rgb = im2double(imread(flowers.tif)); whos rgb Name Size Bytes Class rgb 362x500x3 4344000 double array nrows=size(rgb,1); ncols=size(rgb,2); v = reshape(rgb,nrows*ncols,3); white=reshape(min(v),nrows,ncols); Name Size Bytes Class rgb 362x500x3 4344000 double array v 181000x3 4344000 double array white 362x500 1448000 double array clr(:,:,1)=rgb(:,:,1)-white; clr(:,:,2)=rgb(:,:,2)-white; clr(:,:,3)=rgb(:,:,3)-white; rgb2gray(rgb) min(r,g,b) c = v*v/N [V,D]=eig(c) c = 0.2422 0.1677 0.1468 0.1677 0.1391 0.1237 0.1468 0.1237 0.1508 V = 0.4175 0.5960 0.6860 -0.8520 -0.0058 0.5235 0.3160 -0.8030 0.5053 D = 0.0111 0 0 0 0.0428 0 0 0 0.4783 Correlation matrix Eigenvectors Eigenvalues E = EIG(X) is a vector containing the eigenvalues of a square matrix X. [V,D] = EIG(X) produces a diagonal matrix D of eigenvalues and a full matrix V whose columns are the corresponding eigenvectors so that X*V = V*D. a = reshape(v*V(:,3),nrows,ncols); amax = max(max(a)) imshow(a/amax); imwrite(a/amax,’flowers1.jpg’); amax = 1.7148 ? a = reshape(v*V(:,2),size(a)); ? amax = max(max(a)) ? amin = min(min(a)) ? b = (1-a/abs(amin))/2; ? imshow(b) ? imwrite(b,flowers2.jpg); amax = 0.3912 amin = -0.5912 *