i want to convert a RGB picture to YUV,and the restore it .but when i calculate in MATLAB,i found The picture isn't quite true to the original. here is my Operation in matlab ! here is my code:
%M file
clear all;
clc;
RGB = imread('11111.bmp');
figure;imshow(RGB);
R = RGB(:,:,1);
G = RGB(:,:,2);
B = RGB(:,:,3);
Y0= 0.256788*R + 0.504129*G + 0.097906*B + 16;
U0=-0.148223*R - 0.290993*G + 0.439216*B + 128;
V0= 0.439216*R - 0.367788*G - 0.071427*B + 128;
Y= uint8(min(255,max(0,Y0)));% i want 8 bit YUV
U= uint8(min(255,max(0,U0)));
V= uint8(min(255,max(0,V0)));
RGB1R=Y+1.14 *V;
RGB1G=Y-0.39 *U-0.58*V;
RGB1B=Y+2.03*U;
RGB1=cat(3, RGB1R, RGB1G, RGB1B);
figure; imshow(RGB1);
%M file
it should point out that the image is capsule endoscopy image.you could get it in wwww.gastrolab.net .i want the YUV is 8 bits interger,could anybody help me??
To convert between RGB and YCbCr, MATLAB offers the functions rgb2ycbcr and ycbcr2rgb. YCbCr and YUV are then related by
Y = Y
U = 0.872021 Cb
V = 1.229951 Cr
To convert an RGB image to YUV, you can thus use
RGB = imread('11111.bmp');
YCBCR = rgb2ycbcr(RGB);
YUV(:,:,1) = YCBCR(:,:,1); % Y -> Y
YUV(:,:,2) = 0.872021 * YCBCR(:,:,2); % Cb -> U
YUV(:,:,3) = 1.229951 * YCBCR(:,:,3); % Cr -> V
and to convert it back, use
YCBCR1(:,:,1) = YUV(:,:,1); % Y -> Y
YCBCR1(:,:,2) = YUV(:,:,2) / 0.872021; % U -> Cb
YCBCR1(:,:,3) = YUV(:,:,3) / 1.229951; % V -> Cr
RGB1 = ycbcr2rgb(YCBCR1);
Note that if you are using uint8 as type for RGB, then YCBCR and YUV will be uint8 too and the conversion will be lossy. For a testimage of mine, the maximum difference between RGB and RGB1 was
max(abs(RGB(:)-RGB1(:)))
ans =
4
To have a lossless conversion, you should use the double datatype. Note that MATLAB assumes that for uint8 images the values are between 0 and 255, while for double images, the values are between 0 and 1, so they would need to be scaled.