After the talk with Norm and Joe, I've set up a buffer area around my source image so that the optimization does not sample from the edges. However, I'm still not getting very good results.
So I created some simple test images to see what's going wrong with the optimization. Here are some results:
No comments:
Post a Comment