[ptx] autopanos?

[alexandre jenny]

Let's try some comparison (I'll try to be the most accurate, but sebastian,
if I make some mistake, correct me).

Autopano Alexandre Jenny (That's me ;-))
 + Speed, it's c++.
 + Very low memory usage.
 + Ransac check (removes many false point).
 - Accuracy. Around 5 % of false points still remains.
 - Not open source yet.

Autopano_sift Sebastian
 + Accuracy. It seems to be more accurate.
 + Open source.
 - Speed.
 - Memory usage.

Am I right ?

Bye
  Alexandre 

We should use some test case to make a real comparison ... Hum ...

> -----Message d'origine-----
> De : ptx-bounces at email-lists.org 
> [mailto:ptx-bounces at email-lists.org] De la part de Pablo d'Angelo
> Envoyé : mercredi 5 mai 2004 23:42
> À : ptx at email-lists.org
> Objet : Re: [ptx] autopanos?
> 
> 
> On Wed, 05 May 2004, JD Smith wrote:
> 
> > Can anyone offer a comparison of Alexandre's autopano and 
> Sebastian's 
> > autopano-sift in terms of features, patent/license encumbrances, 
> > memory usage, performance on large (~3-6Mpix) images, underlying 
> > algorithms, etc.?
> 
> Not that I know of. I'd like to do one some time in the 
> future, but I'm too busy with coding right now.
> 
> btw. panosifter, included in the hugin package could be used 
> for this purpose as well, mainly to test the different sift 
> implementations against the reference implementation by D.Lowe.
> 
> The test cases in my gallery might be a start.
> 
> ciao
>   Pablo
>