Calibration between two RGBD sensors who dont have common FOV

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

Calibration between two RGBD sensors who dont have common FOV

shome
I have two primesense RGBD sensors which are fixed back to back and dont have
any common Field of view. I need to find the extrinsics(Rotation and
translation) between the two cameras. I need to calibrate either the two RGB
cameras or two IR cameras in the respective sensors

Options which I could think was by adding intermediate RGBS sensors with
overlapping views. Then try ICP with pointclouds from neighboring pair of
sensors. Then multiply the transformations. However this leads to the loop
closure problem as a result of accumulation of errors induced by ICP.


Anyone can kindly give any ideas?



--
Sent from: http://www.pcl-users.org/
_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users
Reply | Threaded
Open this post in threaded view
|

Re: Calibration between two RGBD sensors who dont have common FOV

Hilarius
You can scan a room and generate a loop closed point cloud using one
system. Afterwards do a shot with both systems simultaneously and
calculate both positions from that based on the room setup. Then you can
calculate your transformation from that.
You can even do multiple shots and estimate best transformation values.
Regards
Felix

Am 10.10.2017 um 13:01 schrieb shome:

> I have two primesense RGBD sensors which are fixed back to back and dont have
> any common Field of view. I need to find the extrinsics(Rotation and
> translation) between the two cameras. I need to calibrate either the two RGB
> cameras or two IR cameras in the respective sensors
>
> Options which I could think was by adding intermediate RGBS sensors with
> overlapping views. Then try ICP with pointclouds from neighboring pair of
> sensors. Then multiply the transformations. However this leads to the loop
> closure problem as a result of accumulation of errors induced by ICP.
>
>
> Anyone can kindly give any ideas?
>
>
>
> --
> Sent from: http://www.pcl-users.org/
> _______________________________________________
> [hidden email] / http://pointclouds.org
> http://pointclouds.org/mailman/listinfo/pcl-users

_______________________________________________
[hidden email] / http://pointclouds.org
http://pointclouds.org/mailman/listinfo/pcl-users