This is the full setup of OpenCV with CUDA and cuDNN support for the Jetson Nano.
For more information see Q-engineering - Install OpenCV Jetson Nano
The default memory (4 GB RAM + 2 GB swap) on your Nano is not enough for a quick build.
In this case the compilation will be done by 1 core, which will take a long time.
You must have more memory allocated to your Nano for the fast 4 core build.
# check your total memory (RAM + swap) first for a fast build. You need at least a total of:
# OpenCV 4.5.4 -> 8.5 GB!
# OpenCV 4.5.3 -> 8.5 GB!
# OpenCV 4.5.2 -> 8.5 GB!
# OpenCV 4.5.1 -> 6.5 GB
# OpenCV 4.5.0 -> 6.5 GB
# if not, enlarge your swap space as explained in the guide,
# or only 1 core is used for the compilation.
$ free -m
$ wget https://github.com/Qengineering/Install-OpenCV-Jetson-Nano/raw/main/OpenCV-4-5-x.sh
$ sudo chmod 755 ./OpenCV-4-5-x.sh
$ ./OpenCV-4-5-x.sh
👉 Don't forget to reset your swap memory afterwards.
If you want to beautify OpenCV with the Qt5 GUI you need to
- $ sudo apt-get install qt5-default
- Set the -D WITH_QT=ON \ (± line 62) in the script
before running the script on your Nano
OpenCV will be installed to the /usr directory, all files will be copied to following locations:
/usr/bin- executable files/usr/lib/aarch64-linux-gnu- libraries (.so)/usr/lib/aarch64-linux-gnu/cmake/opencv4- cmake package/usr/include/opencv4- headers/usr/share/opencv4- other files (e.g. trained cascades in XML format)
