Question Regarding validation with own data and the variable "smoothed_object_pc"

[gtan039]

Hello, can I ask, how is the "smoothed_object_pc" variable derived? I am able to get the RGB image data, depth image, and the intrinsic matrix of the camera, but I am not sure about "smoothed_object_pc" as well as "base_to_camera_rt

[arsalan-mousavian]

It's explained here: https://github.com/NVlabs/6dof-graspnet/blob/master/demo/main.py#L127-L128

base_to_camera_rt: is the relative transformation between camera and robot base which is not used in the example.

[ggzzzzz628]

It's explained here: https://github.com/NVlabs/6dof-graspnet/blob/master/demo/main.py#L127-L128

base_to_camera_rt: is the relative transformation between camera and robot base which is not used in the example.

So how does it work? Can you be more specific? thank you

[ggzzzzz628]

It's explained here: https://github.com/NVlabs/6dof-graspnet/blob/master/demo/main.py#L127-L128

base_to_camera_rt: is the relative transformation between camera and robot base which is not used in the example.

I now want to generate the smoothed_object_PC format from Realsense. How do I do that? Can you help me? Thank you

[ggzzzzz628]

It's explained here: https://github.com/NVlabs/6dof-graspnet/blob/master/demo/main.py#L127-L128 base_to_camera_rt: is the relative transformation between camera and robot base which is not used in the example.

I now want to generate the smoothed_object_PC format from Realsense. How do I do that? Can you help me? Thank you

I understand it's 10 frames of point cloud data, right?

[patselle]

Hello, I am also interested into how to perfom this step.

Can you please share this code snipped in order to do so?

Best thanks I hope it is fine for you