let image x be an original (non warped) image. let I1 be I1' having suffered some projecive transofmation and possibly some noise. We have computed SIFT descriptors for both images.
can you explain how RANSAC (concretely) is used to find the homography that aligns image x and y? more concretely, I understand we give to the RANSAC function the keypoints(I1) and keypoints(I1'). My understanding (from various papers/tutorials i ve read) is that it will take the postions of keypoints in I1 that i note (x,y) and the positions of keypoints I1' (x',y') i.e. RANSAC((x,y),(x',y')) and then take randomly paris of (x',y') (x,y) and then copute the homography by trying to solve for each xi'yi' : [x'i,y'i]=H*[xi,yi] then how to estimate inliers ? I guess by looking at the Euclidean distance bewteen each pairs of points found in the homography....
However I don't really see the 'big picture' ... I don't completey link this process of 'finding homoraphy' and checking inliers etc.
It would be great if someone could summarize that in a clear way?
best regards, O.
P.S. In general, tutorials say that they "find the homography" and then check for inliers - in some way - but they don't explain in a more pragmatic way what really happends.... w.r.t sift descriptors