/taraxl-isaac-package

Primary LanguageC++

STEEReoCAM camera Isaac SDK for libargus

STEEReoCAM is a 2 MP 3D MIPI Stereo camera for NVIDIA® Jetson AGX Xavier/TX2/Nano development kit which supports 2 MP ((2*1600) x 1300) at 30 fps. This MIPI Stereo camera is based on 1/2.9" OV2311 global shutter CMOS sensor from OmniVision.

Prerequisites

  • Jetpack 4.3
  • Isaac sdk 2019.3
  • Nvidia TX2,Nano,Xavier

Installation

  1. Download the BSP package from the following link.

    https://www.dropbox.com/sh/8fdb9wshksnuw0f/AACYV0MdNdpd3tahOfykpbTka?dl=0

  2. Copy the release package into the home directory of the flashed Jetson development kit.

  3. Run the following commands to extract the release package in the Jetson development kit to obtain the binaries.

    tar -xampf e-CAM20_Stereo_CUMI2311_TX2_JETSON_<L4T_Version>_<release_date>_<release_version>.tar.gz
cd e-CAM20_Stereo_CUMI2311_TX2_JETSON_<L4T_Version>_<release_date>_<release_version>

  4. Run the following commands to install the binaries.

    sudo chmod a+x ./install_binaries_<version>.sh
sudo ./install_binaries_<version>.sh

    The above script will reboot the Jetson development kit automatically afterinstalling the binaries successfully.

  5. Download ISAAC SDK Version 2019.3 from the following website:

    https://developer.nvidia.com/isaac/downloads

  6. Extract the release package in a local folder.

  7. git clone https://github.com/econsystems/taraxl-isaac-package.git

  8. Run the install.sh script to deploy taraxl related packages

     cd taraxl-isaac-package
 ./install.sh
 Enter the absolute path of Isaac SDK( Eg : /home/nvidia/isaac): <ISAAC Directory>

  9. Go to your isaac directory and run the below commands to run the samples:

     cd <ISAAC Directory>

    For STEEReoCAM Camera(MIPI):

        ./engine/build/deploy.sh --remote_user nvidia -p //apps/samples/taraxl_camera:taraxl_camera-pkg -d jetpack43 -h <ROBOT IP>
    ./engine/build/deploy.sh --remote_user nvidia -p //apps/samples/taraxl_imu:taraxl_imu-pkg -d jetpack43 -h <ROBOT IP>

Now you can launch the application in the device/robot.