Specifying path of source for nvarguscamerasrc

Accelerated Computing Intelligent Video Analytics DeepStream SDK
,
1

So one can specify a camera using the sensor-id parameter:

For instance for /dev/video3 one can do:
$ gst-launch-1.0 nvarguscamerasrc sensor-id=3 ! 'video/x-raw(memory:NVMM), width=(int)1280, height=(int)1080, format=(string)NV12' ! nvvidconv ! queue ! xvimagesink

Now if my camera’s name is different, how can I get nvarguscamerasrc to point that source?

For instance, if my camera is /dev/mycamera3, what can I do?
Neither:
$ gst-launch-1.0 nvarguscamerasrc '/dev/mycamera3' ! 'video/x-raw(memory:NVMM), width=(int)1280, height=(int)1080, format=(string)NV12' ! nvvidconv ! queue ! xvimagesink

nor
$ gst-launch-1.0 nvarguscamerasrc src='/dev/mycamera3' ! 'video/x-raw(memory:NVMM), width=(int)1280, height=(int)1080, format=(string)NV12' ! nvvidconv ! queue ! xvimagesink

works.

3

Hi,
Is your camera Bayer sensor or YUV sensro? If it is Bayer sensor, you should configure sensor-id. If it is YUV sensor, you may refer to

4

Thanks for the instructions.

My pixels are ‘BG12’
12-bit Bayer BGBG/GRGR

And yes, it’s the only format available.

However this command does not work: gst-launch-1.0 v4l2src device=/dev/video1 ! 'video/x-raw, format=BG12, width=1280, height=1080, framerate=30/1' ! nvvidconv ! 'video/x-raw(memory:NVMM), format=NV12' ! nvoverlaysink

Error: nvbuf_utils: Could not get EGL display connection WARNING: erroneous pipeline: could not link v4l2src0 to nvvconv0, neither element can handle caps video/x-raw, format=(string)BG12, width=(int)1280, height=(int)1080, framerate=(fraction)30/1

Any clue how to get this to work?

5

Hi,
Yours is a Bayer sensor and need to use nvarguscamerasrc. If the camera is from NVIDIA camera partners, you can get support from the vendor. Our partners are listed in:

If you use your own camera, please check Camera Development in document for programming driver/device tree.