This serves as a documentation & reference for the various resources available for the Jetson Orin Nano, Jetson Xavier NX & Jetson Nano development board with regards to Vision and AI based projects.
The documentation/resources related to the older Jetson Nano (B01) / Jetson Xavier NX development boards would be shifted to the
archivedirectory & the README here.
Documentation is tested on Jetpack 6.2 (
R36 (release), REVISION: 4.4), you could check the system version viacat /etc/nv_tegra_release.
The directory is organised as follows:
- AI_ML: AI & Machine Learning related project(s)
- cpp_proj: C++ related projects
- scripts: Installation & setup scripts. Scripts for the older Jetson Nano and Jetson Xavier NX would be shifted to
archivedirectory inside.
- libargus/Argus (For image processing):
/usr/src/jetson_multimedia_api/argus/ - Multimedia API:
/usr/src/jetson_multimedia_api/
Install python package manager pip.
sudo apt install python3-pipEnsure that the CUDA compiler (nvcc) is added to ~/.bashrc, which could be done so via the add_cuda_path.sh script.
cd ~/cv4-jet-nano/scripts
sudo chmod +x add_cuda_path.sh
./add_cuda_path.shThe jetson-containers would be used to run the various types of language models, such as LLM, VLM, etc.
git clone https://github.com/dusty-nv/jetson-containers
bash jetson-containers/install.shRun a SLM model.
jetson-containers run $(autotag nano_llm) \
python3 -m nano_llm.chat --api=mlc \
--model TinyLlama/TinyLlama-1.1B-Chat-v1.0The
dustynv/nano_llm:r36.4.0container is12.7GB.
Configure the camera via the jetson-io.py script for the Jetson Orin Nano.
sudo /opt/nvidia/jetson-io/jetson-io.pyThe terminal show display the menu. Select Configure Jetson 24pin CSI Connector.
=================== Jetson Expansion Header Tool ===================
| |
| |
| Select one of the following: |
| |
| Configure Jetson 40pin Header |
| Configure Jetson 24pin CSI Connector |
| Configure Jetson M.2 Key E Slot |
| Exit |
| |
|====================================================================|Select Configure for compatible hardware.
=================== Jetson Expansion Header Tool ===================
| |
| |
| 3.3V ( 1) .. ( 2) i2c3 |
| i2c3 ( 3) .. ( 4) GND |
| GND ( 7) .. ( 8) NA |
| NA ( 9) .. ( 10) GND |
| GND ( 13) .. ( 14) NA |
| NA ( 15) .. ( 16) GND |
| GND ( 19) .. ( 20) NA |
| NA ( 21) .. ( 22) GND |
| GND ( 23) .. ( 24) GND |
| |
| |
| Jetson 24pin CSI Connector: |
| |
| Configure for compatible hardware |
| Back |
|====================================================================|After that, select the camera model that you would be using.
For example, if you are using a single IMX219 camera, select the
IMX219-Aoption, which configures CAM0 forIMX219-Acamera.
=================== Jetson Expansion Header Tool ===================
| |
| |
| Select one of the following options: |
| |
| Camera IMX219 Dual |
| Camera IMX219-A |
| Camera IMX219-A and IMX477-C |
| Camera IMX219-C |
| Camera IMX477 Dual |
| Camera IMX477 Dual 4 lane |
| Camera IMX477-A |
| Camera IMX477-A and IMX219-C |
| Camera IMX477-C |
| |
| Back |
| |
| |
|====================================================================|Select Save pin changes.
=================== Jetson Expansion Header Tool ===================
| |
| |
| 3.3V ( 1) .. ( 2) i2c3 |
| i2c3 ( 3) .. ( 4) GND |
| GND ( 7) .. ( 8) NA |
| NA ( 9) .. ( 10) GND |
| GND ( 13) .. ( 14) NA |
| NA ( 15) .. ( 16) GND |
| GND ( 19) .. ( 20) NA |
| NA ( 21) .. ( 22) GND |
| GND ( 23) .. ( 24) GND |
| |
| |
| Jetson 24pin CSI Connector: |
| |
| Save pin changes |
| Discard pin changes |
|====================================================================|Reboot the Jetson to use the camera!
References
- https://forums.developer.nvidia.com/t/help-with-imx219-cameras/290628
- https://forums.developer.nvidia.com/t/no-cameras-available/320541/2
- https://docs.arducam.com/Nvidia-Jetson-Camera/Application-note/Jetson-io/
Use the nvgstcapture-1.0 command to check whether the camera is working.
nvgstcapture-1.0
nvgstcapture-1.0 --orientation 2 # Rotate image output by 180 degreesReferences
Install jtop via pip.
sudo pip install -U jetson-statsReboot the jetson to use jtop.
jtopInstall Firefox browser via flatpak, as there may be issues installing via snap.
cd ~/cv4-jet-nano/scripts
sudo chmod +x install_firefox_flatpak.sh
./install_firefox_flatpak.shReferences
- https://jetsonhacks.com/2025/07/12/why-chromium-suddenly-broke-on-jetson-orin-and-how-to-bring-it-back/
- https://forums.developer.nvidia.com/t/neither-chromium-nor-firefox-work-with-my-jetson-orin-nano/338669
Access docker without running sudo, by adding user to the docker group.
cd ~/cv4-jet-nano/scripts
sudo chmod +x add_docker_group.sh
./add_docker_group.shCheck if Jetpack is installed via dpkg -l | grep nvidia-jetpack. If it returns nothing, it means that the Jetpack installation is missing.
Install nvidia-jetpack.
sudo apt update
sudo apt install -y nvidia-jetpackAfter the installation, run dpkg -l | grep nvidia-jetpack to check the installation. It should print something similar as shown below.
ii nvidia-jetpack 6.2.1+b38 arm64 NVIDIA Jetpack Meta Package
ii nvidia-jetpack-dev 6.2.1+b38 arm64 NVIDIA Jetpack dev Meta Package
ii nvidia-jetpack-runtime 6.2.1+b38 arm64 NVIDIA Jetpack runtime Meta PackageReferences
jtop might not be updated to display the latest Jetpack version, we could use a patch from jetsonhacks to fix this issue for now.
git clone https://github.com/jetsonhacks/jetson-jtop-patch.git
cd jetson-jtop-patch
chmod +x apply_jtop_fix.sh
./apply_jtop_fix.shReboot after the patch has been applied.
sudo rebootThe Jetpack version should be shown when running jtop again.
jtopReferences
- Main: https://github.com/jetsonhacks/jetson-jtop-patch
- Alternative: https://github.com/cyaninfinite/jetson-jtop-patch
To enter into terminal session inside the GUI (Gnome Desktop).
sudo init 3To return to GUI:
sudo init 5For documentation/testing of the Jetson Nano B01 / Jetson Xavier NX, checkout the README here.
Licensed under the MIT License.