在醫學上測量腦部的儀器逐漸發達，而腦電波圖(EEG)屬於非侵入性的測量儀器，且相較於其他裝置腦電波測量裝置較為便宜，因此，近年來有許多學者投入腦機介面的開發，不只在醫學上的應用，也可提高人類生活的娛樂性。通常會使用無線可攜的EEG裝置來提高方便性，而為了降低可攜裝置電池的消耗以及通訊傳輸頻寬的需求，我們提出了壓縮感知式眼動偵測系統。經由訊號的分析，對腦電波訊號做有效的前置處理，而利用隱馬可夫模型(HMM)將眼動的特徵擷取出來。除此之外，在前端加入了壓縮感測(Compressive Sensing)演算法將訊號壓縮來提高可攜帶性。而為了降低運算複雜度以及提高系統效能，我們改良了系統架構，將前置處理的運算合併在壓縮演算法中並最佳化演算法的參數，讓系統達到良好的效能。在過程中，我們在壓縮感測演算法中發現感測矩陣的選擇，可以大幅提升重建效能。本篇論文針對運算複雜度較高的演算法以硬體實現，由於腦波的取樣率不高，所以我們以硬體成本為考量，採用處理器的架構來設計重建演算法，並以系統驗證的方式在FPGA上進行驗證。最後，我們將會建立一個完整的腦機介面系統，即時的將腦波擷取與辨識，由辨識結果控制電腦的行為，使配戴腦電波儀器者，可直接經由眼動(上下左右)控制電腦進行其他應用。

Due to the rapid evolvement of current medical equipments, electroencephalography (EEG) signal measurement becomes more accessible and cheaper for users. Thus, the development of brain-computer interface (BCI) becomes a hot research topic in recent years. BCI is commonly used for medical usage or entertainment applications. For EEG signal acquisition, wireless portable devices are preferred due to their convenience. In the wireless scenario, power and bandwidth become important issues. Therefore, in this thesis we propose the compressive sensing aided EEG classification system to reduce power consumption and communications bandwidth for portable devices.
We apply proper preprocessing for EEG signals to acquire features, and then use hidden Markov model (HMM) to do eye movement direction classification. In addition, we added compressive sensing in the front end to improve transmission efficiency and merge the pre-processing operation into the compression algorithm to reduce the computational complexity. We also do system optimization to achieve better detection performance. A special sensing matrix is developed to achieve better detection performance and lower computation cost. Since the EEG sampling rate is not high, we use a processor-based hardware architecture to design the reconstruction algorithm for compressive sensing to save hardware cost. The hardware blocks are verified on an FPGA board and incorporated with a PC to form a real time BCI system. With the proposed system, users with a EEG headset can control the arrow keys on the computer by moving their eyeballs toward directions of left/right and up/down.

1 Introduction
1.1 Background
1.2 Research Motivation and Previous Works
1.3 EEG Signal and Eye Movement
1.3.1 EEG Signal Acquisition
1.3.2 Event-Related Potential of Eye Movement Event
1.4 Thesis Organization
2 Proposed Eye Movement Direction Classification System
2.1 Compressive Sensing
2.1.1 Signal Recovery via Orthogonal Matching Pursuit
2.2 Preprocessing of EEG Signal
2.2.1 Independent Component Analysis
2.2.2 Extended Moving Difference and Frequency Domain Analysis
2.3 Classification Algorithm
2.3.1 Hidden Markov Model
2.3.2 Viterbi Decoding
2.4 System Optimization
3 System Optimization
3.1 Optimization of System Parameters
3.1.1 Parameter Optimization for Compressive Sensing
3.1.2 Optimizing Hidden Markov Model Parameters
3.2 Simulation Results
4 Architecture Design
4.1 Hardware of Orthogonal Matching Pursuit
4.1.1 Memory Controller
4.2 Hardware of the Viterbi Decoder
4.3 Fixed-point Simulation
5 Implementation Results
5.1 Design and Implementation of the Brain-Computer Interface
5.2 FPGA Verification
5.3 Real-Time Implementation
6 Conclusion

[1] J. J. Vidal, “Toward direct brain-computer communication,” Annual Review of Biophysics and Bioengineering, vol. 2, pp. 157–180, June 1973.
[2] A. Funase, T. Yagi, Y. Kuno, and Y. Uchikawa, “Prediction of eye movements from EEG,” in International Conference on Neural Information Processing, vol. 3, 1999, pp. 1127–1131.
[3] A. Funase, T. Hashimoto, T. Yagi, A. K. Barros, A. Cichocki, and I. Takumi, “Research for estimating direction of saccadic eye movements by single trial processing,” in IEEE Engineering in Medicine and Biology Society, 2007, pp. 4723–4726.
[4] T. Ito, T. Shinji, H. Sumiya, and M. Baba, “Eye movement-related EEG potential pattern recognition for real-time bmi,” in Proceedings of SICE Annual Conference, 2010, pp. 1055–1059.
[5] S. S. Gupta, S. Soman, P. G. Raj, R. Prakash, S. Sailaja, and R. Borgohain, “Detecting eye movements in EEG for controlling devices,” in IEEE Conference on Computational Intelligence and Cybernetics, 2012, pp. 69–73.
[6] A. Cohen, “Hidden markov models in biomedical signal processing,” in IEEE Conference Engineering in Medicine and Biology Society, vol. 3, 1998, pp. 1145–1150.
[7] E. J. Cands and M. B. Wakin, “An introduction to compressive sampling,” IEEE Signal Processing Magazine, vol. 25, no. 2, pp. 21–30, 2008.
[8] T.-P. Jung, S. Makeig, M. Westerfield, J. Townsend, E. Courchesne, and T. J. Sejnowski, “Independent component analysis of single-trial event-related potentials,” in Proc. ICA’, Jan. 1999, pp. 173–178.
[9] J. A. Tropp and A. C. Gilbert, “Signal recovery from random measurements via orthogonal matching pursuit,” Information Theory, IEEE Transactions on, vol. 53, no. 12, pp. 4655–4666, dec. 2007.
[10] E. Candes, J. Romberg, and T. Tao, “Stable signal recovery from incomplete and inaccurate measurements,” Comm. Pure and Applied Math, vol. 59, no. 8, pp. 1207–1233, 2005.
[11] S. Chatterjee, K. Hari, P. Handel, and M. Skoglund, “Projection-based atom selection in orthogonal matching pursuit for compressive sensing,” in National Conference on Communications, 2012, pp. 1–5.
[12] A. J. Bell and T. J. Sejnowski, “An information-maximization approach to blind separation and blind deconvolution,” Neural Computation, vol. 7, pp. 1129–1159, 1995.
[13] T. W. Lee, M. Girolami, and T. J. Sejnowski, “Independent component analysis using an extended infomax algorithm for mixed sub-gaussian and super-gaussian sources,” Neural Computation, vol. 11, no. 2, pp. 409–433, 1999.
[14] H. Yu, H. Lu, T. Ouyang, H. Liu, and B. Lu, “Vigilance detection based on sparse representation of EEG,” in Proc. IEEE EMBC, 2010, pp. 2439–2442.
[15] P. Blunsom, “Hidden markov models,” 2004.
[16] G. D. Forney and Jr., “The viterbi algorithm,” Proceedings of the IEEE, vol. 61, no. 3, pp. 268–278, 1973.
[17] A. M. Abdulghani, A. J. Casson, and E. Rodriguez-Villegas, “Quantifying the performance of compressive sensing on scalp EEG signals,” in Proc. ISABEL, 2010, pp. 1–5.
[18] J. A. Tropp, M. B. Wakin, M. F. Duarte, D. Baron, and R. G. Baraniuk, “Random filters for compressive sampling and reconstruction,” in IEEE ICASSP, vol. 3, 2006, p. III.
[19] G. Huang and L. Wang, “High-speed signal reconstruction with orthogonal matching pursuit via matrix inversion bypass,” in IEEE Signal Processing Systems (SiPS), 2012, pp. 191–196.