Image Proximity Measures information

[Andreoli__Carlo]

hello everybody, i'm trying write a software to stich images togheter, i am starting out from Image Proximity Measures. I would like to know if someone can point me out some internet url where i can find more information about crossCorrelation because the description in the documentation is a little too strict and i would give a more depp approach to the problem

thank you in advance

[Ying_H_Intel]

Hi Carlo,

You may findsome useful info regarding crosscorrelationin basicimage processing book or some image processing tools's help documenation, like mathworks.

The discription in IPP reference manual mayinclude too many formula so too strick:).But they are really a implement with the formula,

C:\Program Files (x86)\Intel\Composer XE 2011 SP1\Documentation\en_US\ipp\ipp_manual\index.htm

>>The functions described in this section compute the proximity (similarity) measure between an image and a template (another image). These functions may be used as feature detection functions, as well as the components of more sophisticated techniques.

There are several ways to compute the measure of similarity between two images.
1) One way is to compute the Euclidean distance, or sum of the squared distances (SSD), of an image and a template. The smaller is the value of SSD at a particular pixel, the more similarity exists between the template and the image in the neighborhood of that pixel.

...

2) The other similarity measure is the cross-correlation function: the higher is the cross-correlation at a particular pixel, the more similarity exists between the template and the image in the neighborhood of that pixel.

The cross-correlation R_tx(r,c) between a template and an image at the pixel in row r and column c is computed by the equation :

The cross-correlation function is dependent on the brightness variation across the image. To avoid this dependence, the correlation coefficient function is used instead. It is defined as:

for example,
ippiCrossCorrFull_Norm is for cacluate the cross-correlation valueaccording to the the formula _tx(r,c). move the templateimage (i.e blue box)on source image, cacluate one pixel by one pixel as belowfigure.

Butthe template anchor for matching the image pixel is always at the geometric center of the template.

Regards,
Ying

[SergeyKostrov]

Hi Ying,

...
2) The other similarity measure is the cross-correlation function:
...

Two links to equationsin the section 2) of your post are broken. Could you fix it as soon as you have time? Thanks in advance!

Best regards,
Sergey