Re: -ve value of correlation coefficient
From: Armin Doerry (adoerry_at_yahoo.com)
Date: 11/21/04
- Next message: Martin Leese: "Re: image degradation"
- Previous message: Cheok Yan Cheng: "-ve value of correlation coefficient"
- In reply to: Cheok Yan Cheng: "-ve value of correlation coefficient"
- Next in thread: Cheok Yan Cheng: "Re: -ve value of correlation coefficient"
- Reply: Cheok Yan Cheng: "Re: -ve value of correlation coefficient"
- Messages sorted by: [ date ] [ thread ]
Date: Sun, 21 Nov 2004 10:02:57 -0700
Something that is sometimes done is to take the absolute value of the
correlation coefficient, which maps to [0,1] and scale it to [0,255].
This is because -1 indicates "highly correlated" just like +1, whereas
minimum correlation is at 0.
-- Armin Doerry adoerry@yahoo.com ==================== "Cheok Yan Cheng" <yccheok@gmail.com> wrote in message news:422c6523.0411202317.10a7e08b@posting.google.com... > With the 12.2-8 equation found in page 703 of digital image processing > by Gonzalez and Woods, i obtain the correlation coefficient of two > different images in the range [-1, 1] > > if i try to construct a image to represent the correlation coefficient > of these two image, is it correct for me to normalized the value > ranged [-1, 1] to [0, 255]? > > thank you! > > regards > yan cheng
- Next message: Martin Leese: "Re: image degradation"
- Previous message: Cheok Yan Cheng: "-ve value of correlation coefficient"
- In reply to: Cheok Yan Cheng: "-ve value of correlation coefficient"
- Next in thread: Cheok Yan Cheng: "Re: -ve value of correlation coefficient"
- Reply: Cheok Yan Cheng: "Re: -ve value of correlation coefficient"
- Messages sorted by: [ date ] [ thread ]