Elementary Doubt in color space conversion- Image Processing

Status
Not open for further replies.
A

anandkumarcr

Junior Member level 1
Joined
Sep 9, 2011
Messages
18
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,437
Hi there,

I had an elementary doubt with regards to color space conversion (RGB-YCbCr). It is said in the ITU-R 601 recommendations that a zero level in Cb or Cr is 128. What physical significance is this of ? Could anybody make things clearer to me ?
An example would help.

Thank you
 

I think it is just an assumption.
In the conversion matrices from RGB to YCbCr you can see also smthng 128+(...) like this. "128" as an offset.
On the other hand 128 (1000000) represents -127 as considered signed number.

Just my comments, i am not sure if the real fact is this.
Hope helps
 

Hi emresel,

Thanks for your reply
In that case what does "Cb = -127" signify ? Does it mean chromaticity is same as +127 but luminosity is opposite ? I wanted to know the actual physical significance.
 

In my opinion it has no physical meaning. It is a mapping from analog to digital domain converting also color space.
May be any other expert here can explain, otherwise.
Reason using YCbCr color space is to able to process easily. ie in RGB domain to increase brightness you have to process all color components R, G and B. But in YCbCr you just modify Y component (by the way luminosity is stored in Y component only).
Then you have to convert it to RGB domain again to display it. Because display devices are all driven in RGB (such as CRT and TFT).
I understand that you cannot imagine in your mind Cb and Cr components like others Y or R,G,B.
Is it really necessary at all? Just accept it as it is...
Good luck
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…