Hmmm - even when looking at the sift and ransac functions, values even at both ends of the range fail. Having a look at the pictures, there is very little in there for the program to work on - I was hoping the curly cloud would be enough, as the water is different, and the green surface seems to have changed by a fair amount.
I even tweaked things around in photoshop to see if I could force a match but no joy.
Sorry
Rory
You said "Tweaked", without explanation.
You might edit in some artifacts for the software to focus on. Maybe a couple of spots where the pixels have been inverted or had the brightness changed by a fixed amount, to be reverted after stitching.
Previous Topic
Index
