How can I extract object Pose information using Realsense D435?

[mj000111111]

Hello,

I have realsense D435 depth sensor. I want to perform pick operation using robotic arm using this camera. my object looks like this, actually hanging vertical

How can I use realsense camera to get object coordinates w.r.t camera?

I can get distance from camera to object using realsense SDK.

I read in this forum that using realsense we can detect one point on object and extracting that x,y,z coordinates in world coordinates. but is it enough to pick(isn't it required roration matrix?)? between where is that world coordinates defined?

or

is it that camera Pose of object that we can give it to robot?

or what am I thinking is wrong?

thanks.

[MartyG]

Six degrees of freedom (6Dof) devices such as the RealSense T265 Trackng Camera can provide pose. The D435 and D435i do not by default.

There was recently a user on the RealSense GitHub site who was using a robot arm for bin picking. Their own approach for getting pose with a D435 was an OpenCV SolvePnP algorithm.

https://github.com/IntelRealSense/librealsense/issues/3569#issuecomment-475896784

Another user offered a different approach to getting object pose with D435.

https://github.com/NVlabs/Deep_Object_Pose/issues/10