bad rectification and high rms returned by stereoCalibration

asked 2014-05-23 05:09:09 -0500

beco gravatar image

Hi, i have some problems while computing rectification of stereo pairs: stereoCalibration returns an high rms error and i obtain bad rectification pairs. I tried both my rectification program and the stereo_calib.cpp provided with opencv. They both returns similiar rms errors. Plus, i ran my program with the sample stereo pairs in opencv/sample/cpp and i got correctly rectified images. So i believe the problem is in the way i take stereo picture, is it possible?

I use the stereo camera of an Htc Evo 3D (a 3D smartphone) taking pictures of a chessboard pattern. I tried to change the number and the set of pictures used as input but the littlest stereoCalibration rms i got was around 1.5 and the rectified images are totally wrong.
Is there any "suggested" way to take a set of pictures for calibration? Thanks, Andrea

edit retag flag offensive close merge delete

Comments

UPDATE: if i resize all the calibration images before processing them, i can reduce the rms error a lot but the rectified images gain no improvements. I have completely black rectified pair!

beco gravatar imagebeco ( 2014-05-23 08:38:07 -0500 )edit

it is always a good idea to "calibrate" all the picture parts. I mean the corners. When the checkboard always appears on the middle of the picture (with different orientations and sizes) you might get bad results. The checkboard should also appears on the corners , bottom, Top, left , right

r_asmak gravatar imager_asmak ( 2014-05-31 12:12:29 -0500 )edit

Hi I am also trying to do a meaningfull stereo calibration of the camera pair on EVO 3d but still with no good results. Would you mind sharing the status of your work with me maybe we could help each other?

markotruba gravatar imagemarkotruba ( 2014-07-24 10:17:31 -0500 )edit

@markotruba i know you had already read this answer but i fully agree with this as i explain here i tried to bypass the takePhoto() call but the problems persist.

PS. using my photo app reduce rms error but i don't get any improvements in the rectified images (and so in the disparity maps)

beco gravatar imagebeco ( 2014-08-18 12:26:03 -0500 )edit