EOGERT - EOG Event Recognizer Tool
Eogert is a probabilistic online method for detecting fixations, saccades, and blinks from EOG signal. The beginning of the signal (the first 'TRAIN_SECS' seconds) is dedicated to training. This period should contain at least few random saccades and blinks.
The implementation is written in Matlab and it uses normpdf routine, found in the statistics toolbox; if your don't have it, don't bother to buy it but just make your own with
norm_pdf = exp(-0.5 * ((x - mu)./sigma).^2) ./ (sqrt(2 * pi) * sigma).
The main file is eogert.m which calls EMgauss1D.m. In the current code, the EOG signals are loaded from the TEST001.mat file which must therefore also exist in the same folder and the signals are read "realtime" from the loaded files. Modify the corresponding code to read the signals over some actual stream (such as socket).
There is also an offline version available, named eogert_offline. You must input EOG signals to it. The output is saved into a file. Type "help eogert_offline" for further details.
The method is described in
Toivanen, M., Pettersson, K., Lukander, K. (2015). A probabilistic real-time algorithm for detecting blinks, saccades, and fixations from EOG data. Journal of Eye Movement Research, 8(2):1,1-14.