Dear Researcher,
I would like to understand and enhance the color image contrast. I have somme questions to ask, i will be very appreciated if you answer me.
1. What does mean"high dynamic range" and " high contrast"?
2. Is the image contrast affected by the illumination?
3. To enhance the contrast of color image, i use the Retinex algorithm to decompose the image into illumination and reflectance images. Then, i enhance the contrast from the illumination image.
3.1 Is is correct to enhance the contrast of illumination image.? If so, what kind of images that we can apply this idea?
Thank you.
Best Regards
Khadidja BELATTAR
1. What does mean"high dynamic range" and " high contrast"?
High dynamic range means that the image has an higher range of luminance values (The intensity of the image).
https://en.wikipedia.org/wiki/High-dynamic-range_imaging
High contrast means that the difference between luminance /colors in the image is big.
https://en.wikipedia.org/wiki/Contrast_(vision)
2. Is the image contrast affected by the illumination?
lets denote two pixles in the image P1 = (x1,y1) and P2 = (x2,y2)
Contrast(P1,P2) = abs(luminance(P1)-luminance(P2))
luminance(P1) = Illumination (P1) x Reflectance(P1)
so changing the Illumination(P1) by a different amount than Illumination(P2)
will effect the contrast between the two points.
3.1 Is is correct to enhance the contrast of illumination image.? If so, what kind of images that we can apply this idea
enhancing the contrast of the illumination will enhance the contrast in the image. Is it correct it depends for what application. I think for low contrast images it can have a good effect.
Tank you dear Amitay Nachmani, i would like to apply this approach on dermoscopic skin lesion images.They have non uniform illumination and low contrast.
The attached, you find an example of skin lesion image.
The image that you give me is not "high dynamic range" image. Therefor if the only thing you want is to enhance the contrast look at the following code it is a very basic way to enhance contrast in color images.
%% Load image
I = imread('12.jpg');
% Convert the image to YUV to sperate color information from intensity
YUV = rgb2ycbcr(I);
Y = YUV(:,:,1); % Intensity
% Becaue in the image are areas with very high values i used local
% histogram eqlization
Yeq = adapthisteq(Y);
% Replace Y by Yeq
YUVeq = YUV;
YUVeq(:,:,1) = Yeq;
% Convert back to rgb
Ieq = ycbcr2rgb(YUVeq);
figure
title('Before')
imshow(I);
figure
title('After local histogram eq')
imshow(Ieq);
Khadidja Belattar
Can u pass the Matlab Code - Retinex algorithm to decompose the image into illumination and reflection images.
I want to estimate the illumination in image.
- Badges
- Science topic