Eye movement signal recognition method based on EEG signal

Abstract

The invention discloses an eye movement signal recognition method based on an EEG signal. The method includes the following steps of step (1) obtaining an EEG signal when eyes move and performing datapreprocessing on the EEG signal; step (2) for the processing result of the step (1), by use of a modified SK algorithm, determining irrelevant optimal vectors, selecting a small number of support vectors, and introducing a kernel technique into the algorithm for mapping the vectors to a high-dimensional space so as to achieve the purpose of enabling classification; and step (3) by use of an MDM algorithm, solving the optimal hyperplane for the support vectors obtained in the step (2). The method is very reliably by use of the EEG signal for recognition and classification of an eye movement way; furthermore, the SK algorithm and the MDM algorithm (SVM) used in the method minimize the structural risk through maximum marginalization in the case of fixed-experience risk, and can make a classifier have satisfactory learning accuracy and stronger promotion ability.

Description

technical field [0001] The invention relates to an eye movement signal recognition method based on electroencephalogram signals, and relates to the field of electroencephalogram signal feature recognition. Background technique [0002] EEG signals are obviously non-stationary signals. Although a lot of work has been done since the detection of EEG signals in the 1920s, there has been no breakthrough for a long time. With the continuous development of signal processing methods, more and more effective analysis methods have been continuously applied in the analysis of EEG signals. [0003] In 2008, Li Yuanqing et al first introduced the semi-supervised learning algorithm into the recognition of electroencephalogram (EEG) signals, which were used to identify two types of motor imagery tasks (dataset Iva) for the right hand and feet. [0004] Long Jinyi et al. proposed a self-training learning algorithm combined with feature extraction in 2010. [0005] Zhu Xiangyang et al. pr...

AI Technical Summary

Problems solved by technology

[0009] Since the EEG signal is a time-varying and non-stationary signal, it has different frequency components at different times and in different states. At present, there is no very good method to accurately characterize the signal and extract the waveform characteristics of the EEG signal transiently. , and due to individual differences, a lot of research work is needed in this area

Images