Jetson nano B01 camera

[jianing0314]

I want to use this code to accelerate OpenCV. The camera I am using is a fisheye camera. Normally, the fisheye image should be a circle, but after acceleration, some areas are lost and become rectangles.

gs_pipeline0 = f"v4l2src device=/dev/video0 io-mode=2 " \
              f"! image/jpeg, width={width}, height={height}, framerate=60/1, format=MJPG " \
              f"! nvv4l2decoder mjpeg=1 " \
              f"! nvvidconv " \
              f"! video/x-raw, format=BGRx " \
              f"! videoconvert " \
              f"! video/x-raw, format=BGR " \
              f"! appsink drop=1"

And I can’t open four cameras at the same time, opening three or more will show insufficient resources.
I hope you can help me, thank you very very much.

[jianing0314]

The camera is a USB camera

[DaneLLL]

Hi,
You may be hitting the constraint:
connected more than two usb cameras problem on deepstream-app (Jetson Nano Dev Kit) - #12 by DaneLLL

Please try with USB3 cameras to achieve 5Gbps bandwidth.

[jianing0314]

Thank you for your answer, but I still have the question above. Why is my fisheye image missing after using the GStreamer pipeline configuration, changing from a circle to a rectangle

[DaneLLL]

Hi,
It looks like the source is in rectangle. The source frames are JPEGs and gets decoded through hardware decoder. No cropping or resizing is applied. You may check setup of the camera source. Probably it does some cropping and then encode to JPEG.

For comparison you may try software decoder jpegdec plugin.