RECON robust estimation

[Keir Mierle]

http://www.cs.unc.edu/~rraguram/papers/0727.pdf

At some point we should probably implement this.

Keir

[Pierre Moulon]

Hi Keir,

Personnaly I think rely on a more statistical method could be better:

http://www.ipol.im/pub/algo/mmm_orsa_homography/

-- 

Regards/Cordialement,

Pierre M

2012/4/30 Keir Mierle <mie...@gmail.com>

http://www.cs.unc.edu/~rraguram/papers/0727.pdf

At some point we should probably implement this.

Keir

--
You received this message because you are subscribed to the Google Groups "libmv-devel" group.
To post to this group, send email to libmv...@googlegroups.com.
To unsubscribe from this group, send email to libmv-devel...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/libmv-devel?hl=en.