Abstract：The aim of this work is to provide a human-computer interaction system for persons with some physical disabilities. It allows subjects to control devices by controlling the gaze of direction the eye. A humancomputer interface which displays targets repeating moving between the center and eight around directions was designed. The system could detecte the directions of the eyes movement by analyzing the horizontal and vertical EOG when the eyes follow on the certain moving targets. First, the EMG and blink interference were removed from the horizontal and vertical EOG using the mathematical morphology processing. Next, the feature vectors of different eye motion directions were extracted by principal component analysis (PCA). At last, the smoothed EOG signals were projected onto the feature vectors to detect the directions of the eye movement which were corresponding to a moving target. Subjects could operate the speller system in accordance with one's wish and input a character. A total of 8 subjects' results were analyzed and the average information rate (ITR) was 25.7 bits/min. The combination of mathematical morphology and principal component analysis methods was feasible, effective and could extract the information about the directions of the eyes movement. Moreover, the information could be transformed into commands to control external devices effectively.
郑敏敏 高小榕*. 基于眼电的字符输入系统研究[J]. 中国生物医学工程学报, 2012, 31(6): 801-806.
ZHENG Min Min GAO Xiao Rong*. Research of Speller System Based on EOG. journal1, 2012, 31(6): 801-806.
［1］吴湛微. 从EOG扫视信号中提取人眼位置信息的研究［D］. 上海：华东理工大学，2004.
［2］Bin Guangyu, Gao Xiaorong, Yan Zheng, et al. An online multichannel SSVEPbased braincomputer interface using a canonical correlation analysis method ［J］, Journal of Neural Engineering, 2009, 6:046002.
［3］郭琛，高小榕. 基于眼电和稳态视觉诱发电位分析的目光跟踪方法［J］. 中国生物医学工程，2012, 31（3）：382-387.
［4］CHEN YuLuen. Application of tilt sensors in human-computer mouse interface for people with disabilities［J］. Neural Systems and Rehabilitation Engineering, IEEE Transactions on, 2001, 9: 289-294.
［5］陈卫东， 李昕，刘俊，等. 基于数学形态学的眼电信号识别及其应用［J］. 浙江大学学报（工学报），2011，45（4）：644-649.
［6］Musa Hakan ASYALI, Seyfullah Halit OGUZ. A morphology based algorithm for baseline wander elimination in ECG records ［C］. Proceedings of the IEEE International Biomedical Engineering Days. Istanbul Turkey: IEEE, 1992:160-164.
［7］张敏，李陶深，钟淑瑛. 基于Matlab的主成分分析方法（PCA）的实现 ［J］. 广西大学学报（自然科学报），2005，30：74-77.
［8］牟华英. 用主成分分析和Fisher准则的脑电信号分类 ［J］. 科学技术与工程，2009,9（22）：6853-6855.
