I am using a NVIDIA Jetson AGX Xavier to run a neural network for lane detection using ROS. I am unable to get input from the camera that I can then publish as a rostopic. When I tried to use VimbaViewer, it said that no Transport layers were found. When I tried to put the ROS drivers avt_vimba_camera it does not compile. Any help would be appreciated. It needs the libpng12.so.0 to run vimba viewer, which is unavailable for Ubuntu 18.04, even though the Jetson page says that this specific camera is supported (Allied Vision Prosilica GT 1290C). In the past I have been able to successfully get the ROS drivers on the amd64 processor but this one is for the arm64. Any help would be appreciated.
May I know where you get the information?
Can you make sure the camera work with L4T first.
Don’t really know why, but this helped me run the Allied Vision Camera on the Jetson:
sudo apt-get purge ’ * libopencv * ’
sudo apt-get remove ros-*
System Settings → Software & Updates → ‘Other Software’ tab → uncheck nvidia related repos
reinstall ROS using ROS website’s tutorial
~/catkin_ws$ source devel/setup.bash
~/catkin_ws$ catkin_make
Use the avt_vimba drivers for ros on github
Then install the transport layer from the allied vision website