I would like to know if there is a way to compare if picture A is "equal" to picture B in high probability of success.
"equal" means that if i have a picture, let's call it picture A, and in the picture we can see some sort of an object, and another picture, let's call it B, and in this picture we can see the same object but in different view and in different environment, the algorithm would then say that the images are equal.
We can assume that the object is in the center of the image in both pictures.
what i was thinking for now is somehow find the important pixels and then compare if they are the same, but i don't know what to do know or where i can find more helpful information abut that.
compare betrween almost identical images 
Post a Reply 
Replies 
Oldest Newest Rating 
0 
Brandon Elliott
· July 20, 2014
I would code something that compares each pixel (assuming the images are the same size) and calculates the percentage that is the same, and if 25% of the pixels are the same, they are the same picture (adjust the 25% to whatever works best)

2 
amit baz
· July 23, 2014
Mathias Frits Rørvik , i know that already but thanks anyway...but, i wonder if there are libraries or algorithms that can help me to solve this problem or guide me.. in my previous works i used openCV for face recognition and more.. but i couldn't figure out if it can help me in one way or another...

 1