Allied Vision Cameras and Vimba SDK quick guide
Allied Vision
Allied Vision company develops digital cameras for computer vision in a variety of sectors like science and research, medical imaging, traffic monitoring, and sports analytics.
The goal of Allied Vision is for the cameras to adapt to the requirements of the application which they achieve designing the cameras to be modular. This allows them to have a flexible and variable camera portfolio with diversity of sensor, lens mounts, filters and more options for their cameras.
Allied vision includes the development of a software for their cameras with GigE, USB, and MIPI CSI-2 interfaces called Vimba SDK. This software provides code examples, a Vimba Viewer application, Vimba’s Image Transform Library, Vimba’s Defective Pixel Manager, Firmware Updater, and Driver Installer.
The company has also develop a GStreamer plugin for the respective software to record images and use the cameras in GStreamer pipelines.
Vimba SDK
Vimba SDK includes scripts for the cameras driver installation and the necessary library to install the GStreamer plugins.
Currently there are two version of the SDK the Vimba SDK and the VimbaX SDK. VimbaX is the new SDK fully GenICam-compliant and supports latest Alvium features and Vmb APIs.
Check out the differences in the following chart:
| Feature | VimbaX | Vimba |
|---|---|---|
| Windows, Linux, ARM, macOS | Yes, yes, yes, yes | Yes, yes, yes, no |
| Supports latest Alvium features | Yes | No |
| Supports legacy non-GenICam compliant features | No | Yes |
| GigE, USB, and CSI-2 support | Yes | Yes |
| IEEE 1394 support | No | Yes |
| Camera Link Config TL | No | Yes |
| C, C++, and Python API | Yes | Yes |
| .NET API | No | Yes |
| Image Transform Library | Yes | Yes |
| Firmware Updater | Yes | Yes |
| Viewer | Yes | Yes |
Follow the below instructions to learn how to install Vimba SDK and their corresponding GStreamer plugin to capture with pipelines.
VimbaX install guide
Download VimbaX SDK in Allied Vision Downloads page
Decompress the VimbaX SDK tarball
tar -xvf VimbaX_Setup-2023-4-Linux_ARM64.tar.gz
Allied vision manage 3 types of cameras, each type has it own script to install the respective drivers.
- GigE camera users: Go to [InstallDir]/VimbaX_x_x/cti/VimbaGigETL_Install.sh
- USB camera users: Go to [InstallDir]/VimbaX_x_x/cti/VimbaUSBTL_Install.sh
- CSI-2 camera users: Go to [InstallDir]/VimbaX_x_x/cti/VimbaCSITL_Install.sh
Run the installation script for your respective camera:
cd VimbaX_2023-4/cti sudo ./VimbaUSBTL_Install.sh
The installation expected output is:
nvidia@nvidia-desktop:VimbaX_2023-4/cti$ sudo ./VimbaUSBTL_Install.sh Registering GENICAM_GENTL64_PATH Registering AVTUSBTL device types Done Please reboot before using the USB transport layer
Gstreamer plugin - vmbsrc
Allied vision develop the vmbsrc plugin to capture from the cameras in a Gstreamer pipeline. You can found the source code on the gst-vmbsrc gst-vmbsrc repository
Steps to install vmbsrc
Download and extract the shared library with the vmbsrc element from the git release.
cd wget https://github.com/alliedvision/gst-vmbsrc/releases/download/1.0.0/gst-vmbsrc-1.0.0-linux-arm64.zip tar -xvf gst-vmbsrc-1.0.0-linux-arm64.zip
Copy the shared library to the gstreamer plugins directory:
sudo cp deploy/lib/libgstvmbsrc.so /usr/lib/$(uname -m)-linux-gnu/gstreamer-1.0/
In order to use the element, the VimbaC shared library needs to be loadable at runtime. In case this is not complete in the VimbaX installation you can add a file to /etc/ld.so.conf.d/ which contains the path to the shared library.
cd /etc/ld.so.conf.d/ sudo su touch vimbax.conf echo /opt/VimbaX_2023-4/api/lib/ > vmbsrc.conf exit cd
Update the ldconfig cache.
sudo ldconfig -v
Check the plugin is installed:
nvidia@nvidia-desktop:~$ gst-inspect-1.0 | grep vmb vmbsrc: vmbsrc: VimbaX GStreamer source
Vimba install guide
Download Vimba SDK in Allied Vision Downloads page
Decompress the Vimba SDK tarball
tar -xzf Vimba_v6.1_ARM64.tgz
Allied vision manage 3 types of cameras, each type has it own script to install the respective drivers.
- GigE camera users: Go to [InstallDir]/Vimba_x_x/VimbaGigETL.
- USB camera users: Go to [InstallDir]/Vimba_x_x/VimbaUSBTL.
- CSI-2 camera users: Go to [InstallDir]/Vimba_x_x/VimbaCSITL.
Run the script for your camera.
cd Vimba_6_1/VimbaUSBTL sudo ./Install.sh
The installation expected output is:
nvidia@nvidia-desktop:Vimba_6_1/VimbaUSBTL$ sudo ./Install.sh Registering GENICAM_GENTL64_PATH Registering VimbaUSBTL device types Done Please reboot before using the USB transport layer
Gstreamer plugin - vimbasrc
Allied vision develop the vimbasrc plugin to capture from the cameras in a Gstreamer pipeline. Follow the below instructions to install the plugin.
Dependencies:
- Cmake >=3.13. To install Cmake follow instructions in Cmake Installation
Steps to install vimbasrc
Clone the plugin sources from vimbasrc repository
cd git clone https://github.com/alliedvision/gst-vimbasrc.git
Export the Vimba SDK code to a variable
export VIMBA_HOME=[InstallDir]/Vimba_6_1/
Go to the vimbasrc directory, make executable the build script and run the script to install the GStreamer plugin
cd gst-vimbasrc chmod +x build.sh ./build.sh
Expected output:
nvidia@nvidia-desktop:~/gst-vimbasrc$ ./build.sh # ============================================================================ cmake -S . -B build -DVIMBA_HOME=$VIMBA_HOME -- The C compiler identification is GNU 7.5.0 -- Check for working C compiler: /usr/bin/cc -- Check for working C compiler: /usr/bin/cc -- works -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Detecting C compile features -- Detecting C compile features - done -- Found PkgConfig: /usr/bin/pkg-config (found version "0.29.1") -- Checking for module 'gstreamer-1.0' -- Found gstreamer-1.0, version 1.14.5 -- Checking for module 'gstreamer-base-1.0' -- Found gstreamer-base-1.0, version 1.14.5 -- Found GSTREAMER_BASE_LIBRARY: /usr/lib/aarch64-linux-gnu/libgstbase-1.0.so -- Found GSTREAMER_BASE_INCLUDE_DIR: /usr/include/gstreamer-1.0 -- Found GStreamer: /usr/lib/aarch64-linux-gnu/libgstreamer-1.0.so -- Checking for module 'glib-2.0' -- Found glib-2.0, version 2.56.4 -- Found GLIB2: /usr/lib/aarch64-linux-gnu/libglib-2.0.so -- Checking for module 'gobject-2.0' -- Found gobject-2.0, version 2.56.4 -- Found GObject: /usr/include/glib-2.0 -- Configuring done -- Generating done -- Build files have been written to: /home/nvidia/gst-vimbasrc/build # ============================================================================ cmake --build build Scanning dependencies of target gstvimbasrc [ 25%] Building C object CMakeFiles/gstvimbasrc.dir/src/gstvimbasrc.c.o [ 50%] Building C object CMakeFiles/gstvimbasrc.dir/src/vimba_helpers.c.o [ 75%] Building C object CMakeFiles/gstvimbasrc.dir/src/pixelformats.c.o [100%] Linking C shared library libgstvimbasrc.so [100%] Built target gstvimbasrc # ============================================================================
To use the plugin export the build directory to the GStreamer plugin path
export GST_PLUGIN_PATH=~/gst-vimbasrc/build/:$GST_PLUGIN_PATH
Also you can copy the vimbasrc binary to the gstreamer plugins directory
sudo cp libgstvimbasrc.so /usr/lib/aarch64-linux-gnu/gstreamer-1.0/
Check plugin is installed:
nvidia@nvidia-desktop:~$ gst-inspect-1.0 | grep vimba vimbasrc: vimbasrc: Vimba GStreamer source
Cmake Installation
To build the GStreamer plugin you need to build the Cmake package.
If the installed version in your system is a lower than 3.13 uninstall and install the require version.
sudo apt purge cmake
Download cmake sources in https://cmake.org/download/.
wget https://github.com/Kitware/CMake/releases/download/v3.13.4/cmake-3.13.4.tar.gz
Decompress tarball with cmake sources
tar -zxvf cmake-3.13.4.tar.gz
Build and install. This build can take several minutes.
sudo ./bootstrap sudo make sudo make install
Check the installation and the package version
cmake --version
GStreamer pipelines
The GStreamer plugins allows allied vision cameras to be available as GStreamer sources. Using the plugin enables to record images, use specify color formats and handle properties as exposure time and gain.
Camera ID
To use the plugins with a respective camera is important to know the cameras ID, use the list_cameras.py script from the Python APIs which list the information of available cameras.
- For VimbaX the Python package is install in the SDK installation.
cd VimbaX_2023-4/api/examples/VmbPy python3 ./list_cameras.py
- For Vimba install the Python package and run the script.
cd Vimba_6_1VimbaPython/Source python3 -m pip install . cd Vimba_6_1VimbaPython/Examples python3 ./list_cameras.py
vmbsrc pipelines examples
Display image:
gst-launch-1.0 vmbsrc camera=DEV_00012C03ADA4 exposuretime=30000 ! queue ! videoconvert ! xvimagesink
The default format is grayscale, in order to have a color image use:
gst-launch-1.0 vmbsrc camera=DEV_00012C03ADA4 exposuretime=30000 ! video/x-raw,format=UYVY ! queue ! videoconvert ! xvimagesink
Stream via RTSP:
gst-launch-1.0 vmbsrc camera=DEV_0gst-launch-1.0 vmbsrc camera=DEV_00012C03ADA4 exposuretime=30000 ! video/x-raw,format=UYVY ! queue ! nvvidconv ! queue ! nvv4l2h264enc insert-sps-pps=true bitrate=7000000 ! video/x-h264, mapping=/stream1 ! queue ! rtspsink service=500000
vimbasrc pipelines examples
Display image:
gst-launch-1.0 vmbsrc camera=DEV_1AB22C023E6B exposuretime=30000 ! queue ! videoconvert ! xvimagesink
The default format is grayscale, in order to have a color image use:
gst-launch-1.0 vmbsrc camera=DEV_1AB22C023E6B exposuretime=30000 ! video/x-raw,format=UYVY ! queue ! videoconvert ! xvimagesink
Save a video file:
gst-launch-1.0 vimbasrc camera=DEV_1AB22C023E6B exposuretime=30000 gain=1000000 ! video/x-raw,format=UYVY ! queue ! nvvidconv ! queue ! nvv4l2h264enc insert-sps-pps=true bitrate=8000000 idrinterval=10 ! mpegtsmux ! filesink location=test.mp4
Save an image file:
gst-launch-1.0 vimbasrc camera=DEV_1AB22C023E6B exposuretime=30000 gain=1000000 num-buffers=1 ! video/x-raw,format=UYVY ! videoconvert ! pngenc ! filesink location=out.png
Stream via RTSP:
gst-launch-1.0 vimbasrc camera=DEV_1AB22C023E6B exposuretime=30000 ! video/x-raw,format=UYVY ! queue ! nvvidconv ! queue ! nvv4l2h264enc insert-sps-pps=true bitrate=8000000 idrinterval=10 ! perf ! video/x-h264, mapping=/stream1 ! rtspsink service=5000
Troubleshooting
IF the GStreamer source is copy into /opt do not run the build script, it will fail with permission errors.
Instead of using the build script, run the following commands:
sudo cmake --build build -DVIMBA_HOME=$VIMBA_HOME sudo cmake --build build
For direct inquiries, please refer to the contact information available on our Contact page. Alternatively, you may complete and submit the form provided at the same link. We will respond to your request at our earliest opportunity.
Links to RidgeRun Resources and RidgeRun Artificial Intelligence Solutions can be found in the footer below.
ARTIFICIAL INTELLIGENCE
