Background:
Assuming two scenes for the same scene from two different points of view. The application of the registration algorithm on them will lead to the creation of a homography matrix, which is the relationship between them. Due to deformation, one of them, using this homography matrix, (theoretically) will lead to two identical images (if not a common area is ignored).
Since no perfection exists, the two images may not be exactly the same, we can find some differences between them, and these differences can be clearly shown by subtracting them.
Example:
In addition, the lighting condition can lead to huge differences in subtraction.
Problem:
I am looking for a metric to evaluate the accuracy of the registration process. This metric should be:
: 0- > 1 , ( , , ...). , , , 0,5, , ( ) . , - "Hello" .
. , , , ( ).
, , , . , . , , , , .
. .
P.S. OpenCV .