Electroencephalogram characteristic extracting method based on DWT and EMD integrating CSP

Abstract

The invention discloses an electroencephalogram characteristic extracting method based on DWT and EMD integrating CSP. The method comprises the steps of using obtained electroencephalograms as a training set and a testing set and preprocessing the electroencephalograms in channels selected in the training set and the testing set respectively; conducting wavelet decomposition on the electroencephalograms after being preprocessed to obtain sub-band signals after wavelet decomposition and selecting sub-band signals within a set frequency range; conducting empirical mode decomposition on the selected sub-band signals to obtain a natural mode function to obtain reconstructed signals; merging IMF components in the sub-band signals into a matrix, conducting public space mode decomposition on thematrix to obtain a space filter and obtaining characteristic vectors through the space filter; utilizing a support vector machine to conduct characteristic classification training on the characteristic vectors and inputting the preprocessed testing set into the trained support vector machine to classify characteristics to obtain a characteristic classification result. The problem of CSP is input many times and frequency domain information is lacked can be effectively solved.

Description

technical field [0001] The invention relates to an EEG feature extraction method based on DWT and EMD fusion CSP, belonging to the technical field of EEG signal processing. Background technique [0002] The traditional motor channel is composed of brain nerves and muscles, the nerves conduct impulses, and the muscles cooperate to complete corresponding actions, while the Brain-Computer Interface (BCI) provides another motor channel that does not rely on traditional sports The channel is directly connected to the external equipment by the brain consciousness, and the movement channel is established, and the external equipment is directly controlled by the human brain consciousness, without nerve conduction and muscle movement, providing a new exercise method for patients with nerve injury or muscle injury, You no longer need to rely on others to take care of you, you can complete the exercise yourself. The development of brain-computer interface technology can not only help ...

AI Technical Summary

Problems solved by technology

[0007] The technical problem to be solved by the present invention is to overcome the deficiencies in the prior art, to provide a method for extracting EEG features based on DWT and EMD fusion CSP, and to solve the problems of CSP multi-input and lack of frequency domain information in the existing EEG signal extraction methods. Inability to remove impurities and aliased signals from the frequency band

Images