الفهرس 
AbstractOnly 14 pages are availabe for public view
 |  | from | 179 |  |  |
| | | from | 179 | | |
Abstract
This thesis introduces a real-time system of Electrooculogram (EOG) recording that can be effectively used with Human-Computer-Interaction (HCI). The HCI using EOG has been a growing area of research in recent years. The EOG is a signal produced in both positive and negative directions by eye movement. The HCI provides communication channels between the human and the external device. Today, EOG is one of the most important biomedical signals for measuring and analyzing the direction of eye movements. The work in this thesis can be divided into four parts. The first part introduces a design and an implementation of an EOG signal acquisition system. The eye movement analysis software for the human computer interaction is based on real-time data. Two-channel EOG signals are interfaced with Arduino and MATLAB/SIMULINK to acquire data that can be used for hardware control such as LEDs, wheelchairs, robot arms, etc. Six basic eye movements including close eye, open eye, left, right, up and down are considered for control of the hardware devices. The second part introduces a hardware and a software of an EOG acquisition system based on ATmega AVR microcontroller to acquire a dataset of eye movements from volunteers. The system is composed of EOG signal acquisition, Ag / AgCl electrodes, analog-to-digital converter through Arduino Mega 2560 board microcontroller unit, trainer board, laptop, keypad, and liquid crystal display. Different volunteers of different ages at different times have been treated with the presented system to obtain data. The third part depends on the dataset of EOG signals to make feature extraction and classification. The dataset of EOG signals was obtained for 27 healthy people: 14 males and 13 females. A total of 54 signals for 27 healthy individuals are obtained for each direction, each lasting 30 seconds. The Bo-Hjorth parameter is adopted for feature extraction from the preprocessed EOG signals. For classification, Decision Tree (DT), K-Nearest Neighbor (KNN), Ensemble Classifier (EC), Kernel Naive Bayes (KNB) and Support Vector Machine (SVM)) are utilized. The obtained results reveal that the best classifiers on horizontal and vertical signals are the Support Vector Machine (SVM), the Cosine KNN and the Ensemble Subspace Discriminant with accuracy levels up to 100%. The fourth part introduces a control of wireless robot arm based on the EOG signal acquisition system. The NRF24L01 - 2.4G wireless transceiver module is used to transfer data from master to slave circuits. The L298N driver circuit is mounted on a special robotic arm. Robotic arm movement is controlled by five basic eye movements. The performance and accuracy of the presented system are really promising due to simplicity and reliability.
Finally:
this trend can help the disabled patients to take decisions for better life quality through wheelchair regulation, rehabilitation aids and other applications by simply moving their eyes via this presented system, by providing possible human interaction with a computer.