Procesiranje in klasifikacija EEG signalov za napredne vmesnike človek-stroj
EEG-based human-machine interfaces offer an alternative means of interaction with the environment that relies solely on interpreting brain activity. They can not only significantly improve the life quality of people with neuromuscular disabilities, but also present a wide range of opportunities for industrial and commercial adoption. This thesis focuses on processing and classification of motor imagery EEG recordings. The used data consisted of three data sets, two of which were recorded within this project. A software framework that supports EEG signal filtering, feature extraction and classification was developed and successfully used. Selected instances of FIR and IIR digital filters were implemented and compared, showing that the latter was more appropriate for the current application. Several feature extraction methods were implemented, including band power, autoregressive modelling, Hjorth parameters and FFT-based features. An LDA-based linear classification method was implemented and tests have shown that it performs best with the band power features. Additionally, an LSTM-based neural classification method was implemented and optimised in terms of architecture shape, learning rate and weight decay parameters. Through optimisation, it was found that this method also performs best with band power features. The implemented classifiers were compared based on the band power feature, using the available data sets recorded with wet and dry electrodes, with monopolar and bipolar montage. The two methods achieved similar performance in terms of prediction accuracy, although the linear classifier was for the given data and training approach found to be favourable due to its robustness and low complexity.
2017
2017-09-11 13:32:32
1060
elektroencefalografija, vmesnik človek-stroj, vmesnik možgani-stroj, zamišljanje gibanja, digitalno filtriranje, izluščanje značilk, klasifikacija
electroencephalography, human-machine interface, brain-machine interface, motor imagery, digital filtering, feature extraction, classification
mb22
Gal
Gorjup
70
Rok
Vrabič
991
Poramate
Manoonpong
994
VisID
16
200035
1041.pdf
15683615
Predstavitvena datoteka
2017-09-11 13:32:41