基于稳态视觉诱发脑电信号的频率特征提取方法研究

The Study of Feature Extraction for Steady State Visual Evoked Potential Based on Electroencephalograph

脑电信号分析与处理是脑-机接口技术的关键环节,视觉诱发电位是脑-机接口技术较为常用的一种方法。采用功率谱估计中的自相关法、Welch法和AR模型法对稳态视觉诱发脑电信号进行频率特征提取,根据Fisher线性分类对3种方法提取的特征量进行分类判别。结果表明,AR模型法提取频率特征量的准确率最高。

EEG analysis and processing is the key part of brain-computer interface technology. In this paper, the steady-state visual evoked EEG, feature extraction of evoked potentials with applying power spectrum estimation like correlation method, Welch method and the AR model method is processed. The three methods for extracting characteristic quantities are classified based on Fisher linear classification. The results show that AR model method to extract the frequency characteristics has the highest accuracy rate.