Brain-computer interface is a interactive system which can connect with the external environment without pe-ripheral nerves and muscles for people. The P300-based brain-computer interface system is a typical application of BCI technology. In this paper,we based on the conventional P300-based brain-computer interface system,designed and im-plemented a P300-based brain-computer interface with familiar face paradigm. We established signal acquisition and pro-cessing model based on familiar face paradigm. After data preprocessing and feature extraction, we used support vector machine ensemble to classify EEG signals. The results showed that P300-based brain-computer interface with familiar face paradigm evoked the Vpp and N170 potentials besides P300 potential. Compared with the conventional P300-based brain-computer interface system which the character input accuracy was 80.6%,the familiar face paradigm character in-put accuracy reached 93.5%. The results proved that P300-based brain-computer interface system with familiar face par-adigm has a good development prospect.%脑机接口系统是一种使大脑能够不依赖于外周神经和肌肉通道，与外部环境进行交互的系统。基于P300的脑机接口字符输入系统是脑机接口技术的一种典型应用。对基于传统范式的P300脑机接口系统进行改进，设计并实现了基于熟悉人脸范式的P300脑机接口字符输入系统。实验建立了基于熟悉人脸的P300脑机接口系统的信号采集与处理模型，对采集到的数据进行预处理、特征提取，并使用集成支持向量机算法对脑电信号进行分类。结果表明，除P300电位外，熟悉人脸范式诱发出了Vpp和N170电位。与以往传统范式80.6%的字符输入正确率相比，基于熟悉人脸范式的字符输入正确率达到93.5%，具有良好的发展前景。
展开▼
