Electroencephalogram signal classification method combining improved EMD algorithm with wavelet packet transformation and CSP algorithm

Abstract

The invention provides an electroencephalogram signal classification method combining an improved EMD algorithm with wavelet packet transformation and a CSP algorithm. The method comprises the following steps: firstly, carrying out wavelet packet transformation on a preprocessed signal; and then, performing empirical mode decomposition on the reconstructed narrow-band signal, screening out an intrinsic mode function concentrated in an imaginary motion frequency band according to a correlation coefficient of each order intrinsic mode function, further performing CSP filtering, and performing feature selection through a support vector machine to obtain a final classification result. According to the method, the EEG signal is subjected to frequency domain filtering by utilizing wavelet packettransformation before empirical mode decomposition, so that the signal-to-noise ratio of the EEG can be effectively improved. The CSP filtering is performed by utilizing the intrinsic mode function after the EMD of the C3 and C4 channels is decomposed, and the frequency domain information of the EMD is added on the basis of the CSP, so that the problem that the CSP lacks the frequency domain information is well solved.

Description

technical field [0001] The invention relates to the field of intelligent information processing, in particular to an improved EMD algorithm combined with wavelet packet transform and CSP algorithm for classification of brain electrical signals. Background technique [0002] Brain Computer Interface (BCI) is a control technology involving many disciplines and knowledge fields. The brain-computer interface is a direct information transmission between the human brain and external devices, and judges people's intentions by collecting and extracting the EEG signals generated by the brain. The development of brain-computer interface technology not only helps paralyzed patients use electronic devices such as computers, neural prostheses, and robotic arms, but also realizes other functions including: motion recovery, communication, environmental control, and even entertainment. [0003] Brain-computer interface technology mainly includes five steps, which are signal acquisition, pr...

AI Technical Summary

Problems solved by technology

At the same time, conventional public spatial pattern decomposition methods require a large number of input channels, lack frequency domain information, and their development is limited.

Images