Eye Blinking Signal Identification and Verification Based on Neural Network and Peak Correlation

Human-computer interaction (HCI) technology is an essential platform for human and electronics. In recent years, with the rapid development of science and technology, which has made significant progress, where the combination of bioelectrical signal and electronics is the future growth trend, such as electroencephalogram (EEG), electromyography (EMG) and electrocardiograph (ECG). In this thesis, aiming at the bioelectric signal and based on neural network and peak correlation, this thesis project establishes eye blinking detection platform and propose the difference between neural network and peak correlation, which rich bioelectric signal detection theory. This thesis project mainly researches electro-oculogram (EOG) acquisition, data pre-processing and analyzing, eigenvalue extraction and eye blinking recognition based on neural network and peak correlation. The main contents are as follows: collect EEG by using brain wave module and extract EOG based on Butterworth filter and Empirical Mode Decomposition (EMD). Receive three different eye blinking signal which include eye blinking with effort, normal eye blinking and eye blinking with head moving based on EOG features. Implement data pro-processing, energy detection, eye blinking duration and feature extraction. Recognize and validate eye blinking by using neural network and peak correlation and draw a comparative analysis conclusion