QtCamWidget

Qt widget to interact with GigE/IP cameras

Dependencies:

• pyqtgraph
• opencv
• Aravis (GigE library)
• pyepics (to control movement and show slits)

Installation

1. We use Aravis (https://github.com/AravisProject/aravis) to get the stream from the GigE cameras. To build it in Ubuntu start with installing system wide packages:

sudo apt install build-essential meson cmake libgstreamer1.0-dev autoconf intltool \
python-gobject-2-dev gobject-introspection gtk-doc-tools libxml2-dev libgirepository1.0-dev \
libglib2.0-dev libgtk-3-dev libnotify-dev libgstreamer-plugins-base1.0-dev gstreamer1.0-plugins-bad \
libusb-1.0-0-dev

Next we download the latest stable version from http://ftp.acc.umu.se/pub/GNOME/sources/aravis/ and extract it. Follow the instructions at the github page to build it with meson, but make sure to change the installation directory by running "meson configure -Dprefix=/usr/lib" BEFORE you build the package with "meson build", otherwise the files will be installed to /usr/local/lib/x86_64-linux-gnu and you'll have to copy them manually. Check the installation running the arv-viewer-0.x (arv-viewer-0.8 at the time I'm writing this), a widget with fake camera in the list should show up.

2. I recommend to use anaconda and set up a dedicated environment to make averything work, at least for the prototyping phase. Anaconda installation is pretty straightforward. Then create a new environment (skip this if you use existing one)

conda create -n bpmview python=3.7
conda activate bpmview

Next install dependencies: (make sure to use correct versions, latest available didn't work for me)

pip install pycairo==1.18
pip install opencv-python==4.1.2.30
pip install pygobject pyqtgraph pyepics pyqt5

We are almost there.

3. Running the script Download the script from this repository, carefully inspect the code and modify all PV names and addresses. I use perspective correction on the BPMs to show real beam image, these parameters will be different in every case. Activate anaconda environment:

conda activate bpmview

Run the script.

python qt_axiscam.py

If import complains on gi.repository or Aravis go to /usr/lib and copy the "girepository-1.0" folder to ~/anaconda3/envs/bpmview/lib

Now you can open the settings and change camera name/adress/paths and set warping by dragging the circles to form the corner points of the image.