1. Connect the camera
Plug the camera into the USB3.0 port.
2. Download ZED SDK (Jetson)
1. NVIDIA Jetson installs operating system Jetpack
2. Download the ZED SDK that matches the system
Go to the downloaded folder:
cd path/to/download/folder
Use the chmod +x command to increase the execution permissions of the installer. Be sure to replace the name with the version you downloaded:
chmod +x ZED_SDK_JP4.3_v3.0.run
run:
./ZED_SDK_JP4.3_v3.0.run
When the installation starts, the software license Software License will be displayed. Press Enter and press q after reading.
During the installation process, you will need to answer some questions about the installation of dependencies, tools, and samples. Press y and Enter.
Because on the Jetson board, CUDA will be automatically installed together with Jetpack, so you can now use the ZED SDK directly.
3. Run ZED Explorer
ZED Explorer enables real-time preview and recording.
Video resolution, aspect ratio and camera parameters can be changed to capture high-resolution pictures and 3D videos.
Location:
/usr/local/zed/tools/ZED Explorer
4. Run ZED Depth Viewer
ZED Depth Viewer can capture and display depth maps and 3D point clouds.
Try different depth modes to choose the best depth quality/performance ratio.
Location:
/usr/local/zed/tools/ZED Depth Viewer
5. Install ZED Python API
Use ZED SDK in Python3 on Linux platform.
To use ZED SDK in Python, you need to install the following dependencies on your system:
ZED SDK
Python3.6+(x64)
Cython 0.26+
Numpy 1.13+
Opencv Python(optional)
PyOpenGL(optional)
Make sure to install Python (x64 version) and the pip package manager. Then install the dependencies via pip in the terminal:
$ python -m pip install cython numpy opencv-python pyopengl
1. Install Python API
The Python script in the ZED SDK installation folder can automatically detect the platform, CUDA and Python version, and download the corresponding precompiled Python API package.
Run the installation script:
$ cd "/usr/local/zed/"
$ python3 get_python_api.py
# The script displays the detected platform versions
Detected platform:
ubuntu18
Python 3.6
CUDA 11.0
ZED SDK 3.5
# Downloads the corresponding whl package
Checking if https://download.stereolabs.com/zedsdk/3.5/ubuntu18/cu110/py36 exists and is available
Found ! Downloading python package into /usr/local/zed/pyzed-3.5-cp36-cp36m-linux_x86_64.whl
# The script installs the Python API and gives instructions on how to do it manually if needed
Processing ./pyzed-3.5-cp36-cp36m-linux_x86_64.whl
Successfully installed pyzed-3.5
To install it later or on a different environment run :
python -m pip install --ignore-installed /usr/local/zed/pyzed-3.5-cp36-cp36m-linux_x86_64.whl
Note: It is also possible to use a virtual environment, make sure to activate it before running the script.
2. Possible problems
Numpy binary incompatibility
Traceback (most recent call last): ... File "__init__.pxd", line 918, in init pyzed.sl ValueError: numpy.ufunc size changed, may indicate binary incompatibility. Expected 216 from C header, got 192 from PyObject
This error usually means Numpy is not installed, run:
# On Jetson (aarch64) cython needs to be installed first since numpy needs to be compiled.
$ python3 -m pip install cython
$ python3 -m pip install numpy
On Jetson (aarch34), cython needs to be installed first because numpy needs to be compiled.