Raspberry Pi V2 CSI camera works, but not with Cheese Ubuntu 18.04 Jet Pack 4.5

I’m trying to modify a Raspberry Pi object detection project to work on Jetson Nano (4G).
Other discussions seem similar but none have a solution that works for me.

However, I’m struggling with verifying if my camera is set up correctly.
Using the Raspberry Pi V2 camera via the CSI port.

gst-launch-1.0 nvarguscamerasrc | nvoverlaysink - this works
Cheese does not work:
cheese_camera_device_get_name: Assertion “CHEESE_IS_CAMERA_DEVICE” (device)" failed

I tried hooking up a USB camera (a no name dash cab set to ‘PC’ mode ) and it will not work at all

I saw a discussion about csi_camera.ipynb module and usb_camera.ipynb module.
However, not clear what is the issue.

JetPack 4.5
modified Tensorflow installation to use older version due to an error using v2
#import tensorflow.compat.v1 as tf
installed opencv

Any tips would be appreciated.

Pi camera V2 has to be run through nvarguscamerasrc. For launching it in OpenCV, please refer to the samples:

1 Like