python-emotiv is an open-source library to acquire data from Emotiv EPOC headset. Although Emotiv recently provided 32-bit SDK for Linux, we have chosen to use the reverse-engineered protocol as the code will be finally deployed on a Raspberry Pi or BeagleBone Black ARM box.
The library uses libusb to detect and access the dongle instead of hidapi. The udev rules create a /dev/emotiv_epoc symlink in /dev tree, if you want to directly read from that node, pass method="direct" when you create your EPOC object.
Parts of the project are inspired from mushu and emokit which is the pioneer of the reverse-engineered protocol.
- pyusb (Version >= 1.0)
- pycrypto
- numpy
- scipy
- matplotlib (For data analysis scripts under utils/)
- BeagleBone Black GPIO (For SSVEP BCI in examples/)
"labstreaminglayer is a system for the unified collection of measurement time series in research experiments and handles both the networking, time-synchronization, (near-) real-time access as well as optionally the centralized collection, viewing and disk recording of the data."
python-emotiv has preliminary support in lsl/ folder to stream Emotiv EEG signals to labstreaminglayer nodes.
utils.py contains a save_as_matlab() function to export the acquired signals as a MATLAB file. This function saves the EEG data according to the FieldTrip specification to ease the process of analysing signals with FieldTrip.
Installing on macOS:
brew install pygobject3 gtk+3
Just run python setup.py install to install the module on your system.
When you run emotiv/epoc.py as a standalone application, it will dump sensor data to the terminal:
Ozan Çağlayan, Galatasaray University, Computer Engineering Dept.