Quantcast

Stereo Calibration - cvFindStereoCorrespondenceBM

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Stereo Calibration - cvFindStereoCorrespondenceBM

webercalixto
Hi guys,

I'm with stereo calibration and rectification. However,
after cvFindStereoCorrespondenceBM I get the image pair, but awkwardly
rotated clockwise.

There is a sample image of my result, using the chessboard pattern (10X7):
http://img543.imageshack.us/img543/9854/rectification2.png

BMState params:
    BMState->preFilterSize=41;
    BMState->preFilterCap=31;
    BMState->SADWindowSize=41;
    BMState->minDisparity=-64;
    BMState->numberOfDisparities=128;
    BMState->textureThreshold=10;
    BMState->uniquenessRatio=15;

Does anyone have clues about how I get this rotated pair?

--
Weber de Souza Calixto
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Stereo Calibration - cvFindStereoCorrespondenceBM

slug
Does this happen for every calibration? It looks a little strange but maybe another calibration will be ok? Other than that I have no idea why it should even do that.

Reiner



________________________________
 From: weber de souza calixto <[hidden email]>
To: opencv <[hidden email]>
Sent: Tuesday, 7 February 2012 12:13 AM
Subject: [OpenCV] Stereo Calibration - cvFindStereoCorrespondenceBM
 

 
Hi guys,

I'm with stereo calibration and rectification. However, after cvFindStereoCorrespondenceBM I get the image pair, but awkwardly rotated clockwise.


There is a sample image of my result, using the chessboard pattern (10X7):
http://img543.imageshack.us/img543/9854/rectification2.png

BMState params:
    BMState->preFilterSize=41;
    BMState->preFilterCap=31;
    BMState->SADWindowSize=41;
    BMState->minDisparity=-64;
    BMState->numberOfDisparities=128;
    BMState->textureThreshold=10;
    BMState->uniquenessRatio=15; 

Does anyone have clues about how I get this rotated pair?
--
Weber de Souza Calixto



 
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Stereo Calibration - cvFindStereoCorrespondenceBM

webercalixto
Well, not every time, but specially when I try to move the chessboard even
if a little.

If I keep the chessboard steady, the problem is almost no visible, but then
the cameras aren't calibrated and lens distortion isn't fixed.

The stereocalibrate2 is run with 9 chessboard images.

Anyone?

2012/2/6 slug <[hidden email]>

> **
>
>
> Does this happen for every calibration? It looks a little strange but
> maybe another calibration will be ok? Other than that I have no idea why it
> should even do that.
>
> Reiner
>
>   ------------------------------
> *From:* weber de souza calixto <[hidden email]>
> *To:* opencv <[hidden email]>
> *Sent:* Tuesday, 7 February 2012 12:13 AM
> *Subject:* [OpenCV] Stereo Calibration - cvFindStereoCorrespondenceBM
>
>
> Hi guys,
>
> I'm with stereo calibration and rectification. However,
> after cvFindStereoCorrespondenceBM I get the image pair, but awkwardly
> rotated clockwise.
>
> There is a sample image of my result, using the chessboard pattern (10X7):
> http://img543.imageshack.us/img543/9854/rectification2.png
>
> BMState params:
>     BMState->preFilterSize=41;
>     BMState->preFilterCap=31;
>     BMState->SADWindowSize=41;
>     BMState->minDisparity=-64;
>     BMState->numberOfDisparities=128;
>     BMState->textureThreshold=10;
>     BMState->uniquenessRatio=15;
>
> Does anyone have clues about how I get this rotated pair?
>
> --
> Weber de Souza Calixto
>
>
>
>
>  
>



--
Weber de Souza Calixto
Loading...