I'm trying to implement SURF with Ransac for fundamental(F) matrix estimation in Matlab. Then I calculated epipoles as null spaces of F matrix. I used OpenSURF from File Exchange and Ransac from Peter's Kovesi page. However when two images are similar the epipoles doesn't become unstable as they should. Can someone provide me Matlab example of these two working?
2
votes
1 Answers
0
votes
If anyone need this information, great number of examples using these two, one can find in Peter Corke's new book!
