摘要
脑机接口技术是一项不依赖人的外周神经和肌肉组织而实现人机交互通信的技术,使人类可以拥有一条不通过肌肉组织与外界交流而实现人机通信及控制的通道。在总结前人研究成果的基础上,采用了小波变换Mallat算法对实验数据进行特征提取,重点研究了时间窗的选取和利用频率范围选取有用信号的问题。通过SVM的分类结果为准确率84.285 6%,精密度85.272 1%和灵敏度92.000 6%。
Brain-computer interface (BCI) technology is an interactive communication technology which is realized without depending on humans' peripheral nerves and muscles. The technology provides human a man-machine communication and con- trol channel without the communication between muscular tissue and outside world. On the basis of the pervious research re- sults, the Mallat algorithm of wavelet transform is used in the feature extraction of the experiment data. The time window selec- tion and useful signal choosing with the frequency range is focused on. The classification results for accuracy of 84.285 6%, pre- cision of 85.272 1% and sensitivity of 92.000 6% were achieved by SVM.
出处
《现代电子技术》
2013年第19期70-72,76,共4页
Modern Electronics Technique