Camera image stops after prolonged use of the camera

Autonomous Machines Jetson & Embedded Systems Jetson Orin Nano
,
1

A phenomenon similar to the question in the link below is occurring.

There are instances where the camera stops when capturing long-duration camera footage.

Environment:
Jetson Orin Nano
Jetpack 5.1.2

Is there anything that is recognized as an issue?

Although it’s not systematic, by trying out several commands, it seems like there are cases where the footage is less likely to stop and cases where it’s more prone to stop.

⭕ Commands that ran for a long time.

  1. Simple video display (ran continuously for about 2 days, stopped manually)
    Execute the following from the terminal:
gst-launch-1.0 nvarguscamerasrc sensor-id=0 ! "video/x-raw(memory:NVMM),width=4032,height=3040,framerate=20/1" ! nvvidconv ! "video/x-raw(memory:NVMM),width=1920,height=1080,framerate=20/1" ! nvvidconv ! xvimagesink
  1. Writing to shared memory including branching (ran continuously for about 20 hours, stopped manually)
    Execute the following from the terminal:
GST_DEBUG=4 gst-launch-1.0 \
nvarguscamerasrc sensor-id=0 \
! video/x-raw(memory:NVMM),width=4032,height=3040,framerate=20/1,format=NV12 \
! nvvidconv flip-method=1 \
! video/x-raw(memory:NVMM),width=3040,height=4032,framerate=20/1 \
! tee name=mytee \
mytee. \
  ! queue \
  ! nvvidconv \
  ! video/x-raw \
  ! shmsink socket-path=/tmp/image_full wait-for-connection=false \
mytee. \
  ! queue \
  ! nvvidconv \
  ! video/x-raw,width=600,height=800 \
  ! shmsink socket-path=/tmp/image_small wait-for-connection=false

❌ Commands that stop within about 1 to 8 hours.

1.Pipeline using Python’s GStreamer API (to shmsink)

nvarguscamerasrc \
      name=nvarguscamerasrc \
      exposuretimerange='15000000 15000000' \
      gainrange='4 4' \
      ispdigitalgainrange='1 1' \
      aelock=true \
      wbmode=2 \
! video/x-raw(memory:NVMM),width=4032,height=3040,framerate={CAMERA_FRAMERATE}/1,format=NV12 \
! nvvidconv flip-method=1 \
! video/x-raw(memory:NVMM),width=3040,height=4032 \
! nvvidconv flip-method=4 \
! video/x-raw(memory:NVMM),width=3040,height=4032 \
! tee name=mytee \
mytee. \
! queue \
! nvvidconv \
! video/x-raw \
! shmsink socket-path=/tmp/image_full wait-for-connection=false \
mytee. \
! queue \
! nvvidconv interpolation-method=Smart \
! video/x-raw(memory:NVMM),width=600,height=800 \
! nvvidconv \
! video/x-raw \
! shmsink socket-path=/tmp/image_small wait-for-connection=false

Whenever it stops, the following error message is always displayed.
(Log level is os.environ[“GST_DEBUG”] = “nvarguscamerasrc:8,*:4”)

CONSUMER: ERROR OCCURRED
0:49:11.878350917  2509      0xc78a700 INFO                 basesrc gstbasesrc.c:2853:gst_base_src_loop:<nvarguscamerasrc> pausing after gst_base_src_get_range() = eos
0:49:11.878603088  2509      0xc78a700 INFO                    task gsttask.c:312:gst_task_func:<nvarguscamerasrc:src> Task going to paused
0:49:11.878700052  2509      0xc78a640 INFO                    task gsttask.c:312:gst_task_func:<queue1:src> Task going to paused
0:49:11.878779256  2509      0xc78a6a0 INFO                    task gsttask.c:312:gst_task_func:<queue0:src> Task going to paused

This message was the same as when the MIPI flat cable was physically disconnected from the Jetson during camera is runnning.

Additionally, in the following environment where long-duration operation was performed several times, it never stopped.

Environment:
Jetson Xavier NX
Jetpack 5.0.2

Any advice would be greatly appreciated.

1 Like
3

Could you verify on argus_camera of MMAPI.

Thanks

5

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.