Hi
I followed the instructions for “Running the container” from jetson-inference/aux-docker.md at master · dusty-nv/jetson-inference · GitHub
After the models are downloaded
I run again, the command: docker/run.sh
I get an error: xhost: unable to open display “”
I have to mention here, am connected to jetson headless through my laptop over ip address
Could you guide me through how i can continue using in headless mode
And follow the tutorial like when i run: video viewer /dev/video0
i should be able to see the camera open in a new window. in headless mode, i can’t see camera output. but just in the terminal i get a message: vide-viewer: captured 575 frames (1280x 720)
Attach screenshot of the error
rajkumarsaswamy:
I get an error: xhost: unable to open display “”
I have to mention here, am connected to jetson headless through my laptop over ip address
Hi @rajkumarsaswamy , since you are running headlessly, you can ignore this warning.
When running headlessly, you will not see a window with the camera video, because the OpenGL window does not work over SSH tunneling.
To view the video remotely, you need to stream the video over RTP from your Jetson to your PC, as shown here:
Also if you get blurry video quality over RTP, try building jetson-inference from source and running outside from container (issue I am looking into)
I think i may have a basic question now. In the following command to transmit an RTP output stream,
video-viewer --bitrate=1000000 csi://0 rtp://:1234
What is remote-ip, what address i should give here?
That should be the IP address or hostname of your PC.
yes, i got it everything working now.
And when i try building jetson-interface from source and running outside the container, the video quality if better
1 Like
