Usb camera fail on deepstream

[anhnguyen.0912a]

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU): Jetson Xavier NX
• DeepStream Version: 6.3
• JetPack Version (valid for Jetson only): 5.1 (L4T 35.2.1)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
I want to use deepstream for my yolov5. Although I have read several docs and the same problem in Deepstream-test5 for usb camera fail - Intelligent Video Analytics / DeepStream SDK - NVIDIA Developer Forums

I still have a green screen (below).
Here is my config
[application]
enable-perf-measurement=1
perf-measurement-interval-sec=5

[tiled-display]
enable=1
rows=1
columns=1
width=1280
height=720
gpu-id=0
nvbuf-memory-type=0

[source0]
enable=1
#Type - 1=CameraV4L2 2=URI 3=MultiURI
type=1
camera-width=640
camera-height=480
camera-fps-n=30
camera-fps-d=1
camera-v4l2-dev-node=0
uri=usb:/dev/video0

[sink0]
enable=1
type=2
sync=0
gpu-id=0
nvbuf-memory-type=0

[osd]
enable=1
gpu-id=0
border-width=5
text-size=15
text-color=1;1;1;1;
text-bg-color=0.3;0.3;0.3;1
font=Serif

However, with the command

 gst-launch-1.0 v4l2src device=/dev/video0 ! "video/x-raw, width=640, height=480,  format=(string)YUY2, framerate=(fraction)30/1" ! nvvidconv ! 'video/x-raw(memory:NVMM),format=I420' ! nv3dsink

Camera will work

So how could I use deepstream to access camera,

[yingliu]

Can you set option uri as “v4l2:///dev/video0”?

[anhnguyen.0912a]

I have seturi=v4l2:///dev/video0 , but it still does not work

[yingliu]

Does pipeline with uridecodebin gst-launch-1.0 uridecodebin uri=v4l2:///dev/video0 **** work in your setup?
It should be similar as the pipeline mentioned in FAQ 28.3.3.

[anhnguyen.0912a]

gst-launch-1.0 uridecodebin uri=v4l2:///dev/video0 ! nvvideoconvert ! 'video/x-raw(memory:NVMM),format=NV12' ! mux.sink_0 nvstreammux name=mux width=640 height=480 batch-size=1 ! fakesink

I try this command but it still not work. Its output is Preformatted text/dvs/git/dirty/git-master_linux/nvutils/nvbufsurftransform/nvbufsurftransform.cpp:549: => Surface type not supported for transformation NVBUF_MEM_SYSTEM

However, the signal light of webcam turn on

[yingliu]

@anhnguyen.0912a

nvvideoconvert

it should be replaced by nvvidconv, but it is relevant to your deepstream-test5 question.

uri is not needed for USB V4L2, can you set “camera-v4l2-dev-node=0” only?
(uri: URI to the encoded stream. The URI can be a file, an HTTP URI, or an RTSP live source. Valid when type=2 or 3. )

[anhnguyen.0912a]

1. When I change to nvvidconv, nothing popup although the camera signal is turned on.
   And the output is like this
   

   

2. When I change type=2 and set uri=v4l2:///dev/video0, the output is like that
   

   

3. When I set type=1 and have enough key such as

camera-v4l2-dev-node=0
camera-width=640
camera-height=480
camera-fps-n=30
camera-fps-d=1

The output is just a green screen:

[anhnguyen.0912a]

In case only set the source as

enable=1
type=1
camera-v4l2-dev-node=0

the output is look like this:

[yingliu]

Does this pipeline works on your host?

 deepstream-app -c /opt/nvidia/deepstream/deepstream/samples/configs/deepstream-app/source1_usb_dec_infer_resnet_int8.txt

This config file should work well for /dev/video0 which has a resolution of 640x48 with 30fps(same as your working pipeline):

[source0]
enable=1
#Type - 1=CameraV4L2 2=URI 3=MultiURI
type=1
camera-width=640
camera-height=480
camera-fps-n=30
camera-fps-d=1
camera-v4l2-dev-node=0

[anhnguyen.0912a]

I have tried and it still doesn’t work well. Just a green screen popup

[yuweiw]

Have you flashed your os when you update the DeepStream to 6.3? You need to use the JetPack 5.1.2. https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_Quickstart.html#id5

[anhnguyen.0912a]

Oh, my bad.
The reason for issue above may be the conflict of version between Deepstream 6.3 and Jetpack 5.1

I have re-installed Deepstream 6.2 ( or could update Jetpack to 5.1.2) and everything has worked.
Thank you very much for dedicated supporting. !!