Problem detecting webcams connected in Jetson AGX Orin 64 GB

Hello!

I'm trying to connect a Jetson AGX Orin 64 GB module with Matlab R2020b in a PC host with Ubuntu 18.04 via ethernet using "jetson" command from "GPU Coder Support Package for NVIDIA GPUs" app:

hwobj = jetson(ipaddress, username, password);

The connection is successfully accomplished but I receive the next warnings for connected webcams:

Checking for CUDA availability on the Target...
Checking for 'nvcc' in the target system path...
Checking for cuDNN library availability on the Target...
Checking for TensorRT library availability on the Target...
Checking for prerequisite libraries is complete.
Gathering hardware details...
    
Checking for third-party library availability on the Target...
Warning: Unable to fetch information about available webcams. 
Warning: Unable to fetch information about available webcams. 

Gathering hardware details is complete.
 Board name         : NVIDIA Jetson
 CUDA Version       : 11.4
 cuDNN Version      : 8.6
 TensorRT Version   : 5.1
 GStreamer Version  : 1.16.3
 V4L2 Version       : 1.18.0-2build1
 SDL Version        : 1.2
 Available Webcams  :  
 Available GPUs     : Orin 

Then I tried to obtain the list of webcams contected to the Orin module by:

 camlist = getCameraList(hwobj)

but I received:

 Index exceeds the number of array elements (3).

Error in nvidiaboard/getCameraList

Error in jetson (line 13)
camlist = getCameraList(hwobj)

This makes me think that the problem could be the command "jetson" or that I have not installed some requirement in the Orin module. In any case, I haven't solved it yet.

I should say thay I can obtain images and videos from the webcam (Bus 001 Device 009: ID 1224:2a25 Generic 4-Port USB 2.0 Hub) connected to the Orin module using "cheese" software and other applications.

Is there a soluion to this problem?

Thanks for your response in anticipation.

Answer 1

[Ramakrishna Mandalapu]

Hi Gibran,

Thanks for the quick turn around and sorry for the delay in reponse. The latest L4T kernel used in Jetpack is creating dummy video nodes, because of which the MATLAB is not able to find the correct webcam and is failing. This issue was resolved in the latest releases of MATLAB (from 22a). Kindly update to the laest version to work with the Jetson hardware seamlessly. Along with this, there are several other capabilities included in the latest releases such as support for GPIO to AGX Orin, RTP UDP data receive over the network, Sensors (Temperature, magneto, pressure, etc).

Thanks

Ramakrishna

Answer 2

[Ramakrishna Mandalapu]

Hi Gibran,

The software is unable to populate the details about webcams. Can you help us with the camera name?

Also, please run the below command on Jetson and send us the output.

$ v4l2-ctl --info -d <camera index for e.g 0 or 1, etc>

[Gibran]

Hello Ramakrishna,

The webcam that I use is "c920 Logitech HD Pro Webcam". The output that I got using the command $v4l2-ctl --info -d 0 was:

orin@ubuntu:~$ v4l2-ctl --info -d 0
Driver Info:
Driver name      : uvcvideo
Card type        : HD Pro Webcam C920
Bus info         : usb-3610000.xhci-4.4
Driver version   : 5.10.120
Capabilities     : 0x84a00001
Video Capture
Metadata Capture
Streaming
Extended Pix Format
Device Capabilities
Device Caps      : 0x04200001
Video Capture
Streaming
Extended Pix Format
Media Driver Info:
Driver name      : uvcvideo
Model            : HD Pro Webcam C920
Serial           : AC7E139F
Bus info         : usb-3610000.xhci-4.4
Media version    : 5.10.120
Hardware revision: 0x00000011 (17)
Driver version   : 5.10.120
Interface Info:
ID               : 0x03000002
Type             : V4L Video
Entity Info:
ID               : 0x00000001 (1)
Name             : HD Pro Webcam C920
Function         : V4L2 I/O
Flags         : default
Pad 0x01000007   : 0: Sink
Link 0x0200001f: from remote pad 0x100000a of entity 'Processing 3': Data, Enabled, Immutable 

and the output for the command $v4l2-ctl --info -d 1 was:

orin@ubuntu:~$ v4l2-ctl --info -d 1
Driver Info:
Driver name      : uvcvideo
Card type        : HD Pro Webcam C920
Bus info         : usb-3610000.xhci-4.4
Driver version   : 5.10.120
Capabilities     : 0x84a00001
Video Capture
Metadata Capture
Streaming
Extended Pix Format
Device Capabilities
Device Caps      : 0x04a00000
Metadata Capture
Streaming
Extended Pix Format
Media Driver Info:
Driver name      : uvcvideo
Model            : HD Pro Webcam C920
Serial           : AC7E139F
Bus info         : usb-3610000.xhci-4.4
Media version    : 5.10.120
Hardware revision: 0x00000011 (17)
Driver version   : 5.10.120
Interface Info:
ID               : 0x03000005
Type             : V4L Video
Entity Info:
ID               : 0x00000004 (4)
Name             : HD Pro Webcam C920
Function         : V4L2 I/O

So, can I work with this webcam for deep learning tasks even though the software doesn't provide device details?

Thanks for your quick answer.