Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
how imadjust method works in matlab
137 views
Skip to first unread message
Vishwasrao
Jul 14, 2009, 2:39:01 AM7/14/09
to
Hello All,
I am new here ,I am using imadjust method for image enhancement.
I want to know how actually it works with intensities for each pixel.
Can anyone guide me to proper direction.
Thanks in advance.
I am new here ,I am using imadjust method for image enhancement.
I want to know how actually it works with intensities for each pixel.
Can anyone guide me to proper direction.
Thanks in advance.
Ashish Uthama
Jul 15, 2009, 2:48:11 PM7/15/09
to
edit imadjust
Single step (debug) through the code.
YI-NENG
Jun 28, 2012, 2:42:15 AM6/28/12
to
For input image, say f, of class uint8:
Function IMADJUST does the folllowing things
1.Normalizaion: im2double(f), that is , dividing all the entries in f by 255.
2.Gamma Correction is done before linear mapping
For example:
(f,[0 1], [0 0.5], 2) will raise all the entries to the power 2 and then
dividing all the entries by 2 (half).
3.Convert back to the original image class
im2uint8()
uint8 [-->double [0 1]-->gamma correction-->linear mapping-->uint8]
where the things done by IMADJUST is expressed in [].
For class double, the input image values must be in [0 1],
otherwise, the values less than 0 map to 0 and the values higher than 1 map to 1
then do gamma correction and linear mapping.
^^---- Just like that ----^^
^^---- Best regards ----^^
By the way I'm from Taiwan.
Function IMADJUST does the folllowing things
1.Normalizaion: im2double(f), that is , dividing all the entries in f by 255.
2.Gamma Correction is done before linear mapping
For example:
(f,[0 1], [0 0.5], 2) will raise all the entries to the power 2 and then
dividing all the entries by 2 (half).
3.Convert back to the original image class
im2uint8()
uint8 [-->double [0 1]-->gamma correction-->linear mapping-->uint8]
where the things done by IMADJUST is expressed in [].
For class double, the input image values must be in [0 1],
otherwise, the values less than 0 map to 0 and the values higher than 1 map to 1
then do gamma correction and linear mapping.
^^---- Just like that ----^^
^^---- Best regards ----^^
By the way I'm from Taiwan.
0 new messages