Problem with distorted point cloud from two single cameras

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Problem with distorted point cloud from two single cameras

manuel.lo
Hello.
Sorry for the question; I'm not an expert.
I have two ELP single cameras; I wrote a note to obtain a stereoscopic camera publishing relative topics.
I calibrated them following http://wiki.ros.org/camera_calibration/Tutorials/StereoCalibration.
After calibration, I execute stereo_image_proc to produce disparity image and point cloud.
I obtained a good disparity image but I have problem with point cloud.
Using cameras in RTAB-Map, I obtain the same distorted cloud I obtained in RViz; I execute RTAB-Map with launch file I obtained following 2.7 Stereo B in http://wiki.ros.org/rtabmap_ros/Tutorials/SetupOnYourRobot, adapting it to my necessity, but I have, for example, this error:
"A large number (973/1000) of stereo correspondences are rejected! Optical flow may have failed because images are not calibrated, the background is too far (no disparity between the images), maximum disparity may be too small (128.000000) or that exposure between left and right images is too different.
[ WARN] (2018-04-24 18:34:53.579) OdometryF2M.cpp:472::computeTransform() Registration failed: "Not enough inliers 0/5 (matches=0) between -1 and 500"
Have you any suggestion to solve problems?
If you need more info, I'm here.
Thanks a lot.
            Manuel
Reply | Threaded
Open this post in threaded view
|

Re: Problem with distorted point cloud from two single cameras

matlabbe
Administrator
Hi,

[image received by email]Calibration seems not good to me, the disparity image should be a lot more dense than that. See example at the top of this page: http://wiki.ros.org/stereo_image_proc
See also https://www.youtube.com/watch?v=2HT4EPNIPfg

cheers,
Mathieu