The Design And Implementation Of Brain Computer Interface Based On Cross Frequency Scanning Technology

Brain-computer interface technology is a brand new way which don’t rely on the brain, peripheral nerve and muscle system to connect the brain with the outside world(computer or other external devices) in information communication and control.Brain computer interface directly change the brain signal into the external devices’ control commands,the out peripheral nerve and muscle channel is no longer needed in the infomation transfer.The ultimate goal of BCI, is to make the patients with loss of motor ability disease, can operate the computer, voice synthesizer, auxiliary tools and equipment for nerve repair.This interface can increase the patient’s ability to act independently, thus improve the patients quality of life, at the same time reduce the burden of the society.In order to solve the problem that we always don’t have enough frequency points in the brian computer interface experiments.In this paper,it puts forward the idea of dual frequency scanning,namely in a select target cross scanning using two frequency stimulation to obtain the steady-state visual evoked potential, this method can greatly improve the utilization rate of frequency resources, and double frequency scan generate secondary frequency can improve the accuracy of target recognition system, which is likely to build complex brain-computer interface system.According to the demand we designed and implemented dual frequency scanning stimulator of 4 * 4 matrix based on the ARM.And we discussed the selection and collocation of stimulator frequency points.Finally realized the communication of stimulator and PC through serial port, and write a serial communication interface, which is convenient to use.This article also introduces how to extract the eeg signals,to make it a useful signal for us.Given the extreme importance of feature extraction efficiency, the author realized the feature extraction algorithm of this system, use first DAUB4 filtering algorithm for the brain electrical signal preprocessing, reoccupy real Fourier transform algorithm to extract characteristic frequency, and there are described in detail in the paper, especially to the real value of Fourier transform algorithm, this paper in detail analyzes the butterfly operation, binary inversion, such as triangle recursive step from a mathematical formula to algorithm code generation process. In the BCI experiment, eeg data quantity is big and speed is fast. In order to accurately capture and process these data, the paper also implements including the management of the cache area, data of order preserving, and processing the mutually exclusive access to the data between threads, etc.At the end of the paper we designed a brain-computer interface experiment,invited 10 students ot use the BCI system to put in a string which has 30 letters.Through the statistics of the success rate of 10 students.We found that most of the students’ success rate is over 90%, the overall success rate has reached 88.6%.The expriment results show that the BCI system and the BCI expriment are both reasonable.