While reading the wikipedia article on YUV color modulation (which is used in PAL color tv receivers); I visited this diagram:

Wikipedia screenshot (Wikimedia);

(and according to the image description there; i.e.

"An image along with its Y′, U, and V components respectively";

the 3rd one is U-component and 4th one is the V-component of the 1st image. Well.

But there; the component U has been described as

U = approx 0.492 (B - Y') ... (where Y is Luma).

So, if the 3rd image show only U; i.e. 0.492 (B - Y') or 0.492 X (Blue-(0.299R+ 0.587G + 0.114B)) ; then in the ultimate result; there would be equal visual reduction in R and G; and B would be the most prominent.

(As well; they've provided a formula too;
U= -0.147R -0.289G +0.436B ; that shows a positive value for only blue, and a negative values of both R and G)

If someone try to simulate it as visible picture; the picture would contain only-blue (which is non-negative).

But surprisingly, in image 3, there is high portion of green too ; i.e. green is behaving in similar, positive way) ; though there I could not find any trace of red component with ms-paint Eyedropper tool.

labelled as U but doubt

(My eyedropper samples on downloaded original-file ranges from R=0, G=83, B=172 to R=0, G=191, B= 64; and in all samples R=0; on Paint's 255 scale)

So; my question is; is it a wrong picture of U component?

Even if I consider they made a negative copy of U (i.e. gave a -U); still we would obtain Red and Green together as positive; and blue as negative.

U= -0.147R -0.289G +0.436B

= > -U = +0.147R +0.289G -0.436B

But the image shows G and B; not R and G.

In the same way; the 4 th image (V or approx. 0.877(R-Y')) also seems to me showing anomaly from the definitions.

Given equations;

V= +0.615R -0.515G -0.100B

So the positive image would only Red color. But similar to 3rd image; it show proportions of Red and Green; not Red-only.

Labelled as V but doubted

(ms-paint eyedropper samples from original image shown R=147, G=81, B=0 to R=102, G=153, B=0; and in all sampled dots B=0)

If it was a negative copy of V-component; Then we should get positive values of G and B. Not R and G.

So; if these 2 images are correct examples of U and V; what thing I missing about the concept of U and V? what is the basis behind that "extra green" in both the image 3 and 4?

in simple words;

If from a blue image (Component B) if I snatch away a colorless component (Y') then how I get a blue-green picture and not an only-blue picture? or why not blue red?

And same way, if I snatch away a colorless component (Y') from a Red image (Component R); then how I can obtain a Red-Green image and not a only-Red image? or why not blue-red?

  • 4
    \$\begingroup\$ The U and V images should have been displayed in monochrome, just as the Y image was. If a coloured image is a 'vector' containing R, G and B components, then Y, U and V are scalars, they do not have and are not representable as coloured images. \$\endgroup\$ – Neil_UK Dec 24 '16 at 14:53
  • \$\begingroup\$ @user:16880 (David) inviting you to visit this question. \$\endgroup\$ – Always Confused Dec 24 '16 at 18:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.