Identity recognition method based on electroencephalogram signal class multispectral image sequence
A technology of multispectral images and EEG signals, applied in the field of identity recognition based on motor imagery EEG signals to extract multispectral images, can solve the problem of ignoring internal connections, only focusing on time domain features or frequency domain features, and affecting model robustness and performance issues, to achieve the effect of improving accuracy and reliability, good robustness and universality, and improving classification performance
Pending Publication Date: 2022-03-04
HANGZHOU DIANZI UNIV
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, in current research, we often only focus on time domain features or frequency domain features, ignoring the internal relationship between features in different domains, which greatly affects the robustness and performance of the model in practical application.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0032] The method of the present invention will be described in detail below in conjunction with the accompanying drawings, but the embodiments of the present invention are not limited thereto.
[0033] Such as Figure 1~4 Shown, a kind of EEG signal identification method based on multispectral image, comprises the following steps:
[0034] Step 1. Design a test process in which four test pictures and four all-black transition pictures are alternately arranged in a cycle. Each test picture shows the duration of t1, and each transition picture shows the duration of t2; the four test pictures correspond to the Four different motor imagery instructions of the test; in this embodiment, the four test pictures in each cycle use arrows in different directions to represent the motor imagery of the left hand, right hand, tongue, and feet respectively, and the order in which the four test pictures appear is Randomly, the duration of a test cycle is 4(t1+t2); each subject tests N imagin...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an identity recognition method based on an electroencephalogram signal class multispectral image sequence, which comprises the following steps of: designing an electroencephalogram signal acquisition experiment scheme based on motor imagery, and inducing corresponding motor imagery electroencephalogram signals by using arrows in different directions in a test picture; a band-pass filter is used for eliminating ocular artifacts and power frequency interference in original electroencephalogram signals, then a windowing average method is used for obtaining the average power of a specified rhythm at each moment, a multispectral image sequence is generated in combination with two-dimensional distribution of brain electrodes, and the obtained image sequence is regarded as samples to be fed into a deep learning model in batches for identity recognition. According to the method, the motor imagery electroencephalogram signals are selected for identity recognition, so that the feasibility is improved. According to the method, the electroencephalogram signals are used for constructing a multispectral image sequence as a sample, the time-frequency domain features and the spatial domain features of the electroencephalogram signals are fully utilized, and the internal relation of the features is learned by using the deep learning model, so that the model performance is improved, and the robustness is higher.
Description
technical field [0001] The invention relates to the technical field of EEG feature recognition, in particular to an identity recognition method for extracting multispectral images based on motor imagery EEG signals. Background technique [0002] In recent years, more and more researchers have paid more and more attention to the identification technology based on EEG signals, and a lot of scientific research has been carried out. There are many ways to conduct identification and authentication research on EEG signals. According to whether to give the subject a specific stimulus to collect the subject's brain signal for identity authentication, it can be divided into task-based brainprint recognition and task-independent brainprint. Among them, task-based brainprint recognition can be roughly divided into resting-state potential-based brainprint recognition, motor imagery-based brainprint recognition, event-related potential-based brainprint recognition, and visually evoked br...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/044G06F2218/04G06F2218/12G06F18/253
Inventor 孔万增刘国文刘栋军郭继伟崔岂铨
Owner HANGZHOU DIANZI UNIV