Eog

1. Outline
• Objectives
• Introduction
• Literature Review
• Methodology
• Future Work
• Conclusion
• References
ELECTRONICS AND COMMUNICATION ENGINEERING 1

2. Objectives
Acquiring the Electrooculography (EOG) signals
Processing these signals
Extracting the signal features
Classifying the features previously extracted
ELECTRONICS AND COMMUNICATION ENGINEERING 2

3. Introduction
Ø EOG (Electro-Oculogram) signal is a widely used to
detect activities or movements of human eye.
Ø EOG signals uses - control signal for Human-machine
interfaces plays a central role in the understanding,
characterization and classification.
Ø EOG classification system capable of accurately and
consistently classifying eye movements.
ELECTRONICS AND COMMUNICATION ENGINEERING 3

4. Literature Review
ELECTRONICS AND COMMUNICATION ENGINEERING 4
Study Movements Acquisition Processing Method Accuracy
Qi et al. [1] Up, Down, Left, Right Commercial - offline 70%
Guo et al[2] Up, Down, Blink Commercial Laptop online 84%
Kherlopian
et al. [3]
Left, Right, Center Commercial Laptop online 88%
Wu et al.[4] Up, Down, Left, Right,
Up-Right, Up-Left,
Down-Right, Down-Left
Self-designed Laptop +self
designed
online 88.59%
Heo et al. [5] Up, Down, Left, Right,
Blink
Self-designed Laptop
+ self designed
online 91.25%
Erkaymaz et al.
[6]
Up, Down, Left, Right,
Blink, Tic
Commercial Laptop Offline 93.82%
Merino et al
[7]
Up, Down, Left, Right,
Blink, Tic
Commercial Laptop Online 94.11%

5. Workflow
ELECTRONICS AND COMMUNICATION ENGINEERING 5
Data
Acquisition
Low pass
Filter
Standardize
the data
Feature
extraction Classification
Figure 1: Block diagram of methodology

6. Methodology (1/6)
Data Acquisition :
Ø 8 directional eye movements are proposed : up, down, left,
right, up-right, up-left, down-right and down-left
Ø EOG signals are recorded from 2 EOG channels :
vertical(V) and horizontal(H).
Ø A low pass filter was used because the artifacts, arising
from electrodes or head movements and illumination
changes, appear in the high frequencies.
Ø EOG signal information is contained mainly in low
frequencies
ELECTRONICS AND COMMUNICATION ENGINEERING 6
Figure 2: EOG electrodes
placement

7. Methodology (2/6)
Signal Processing :
 The last step in pre-processing was to standardize the data. This is done to remove the
baseline of EOG signals. The standardization was done using the following formula:
i = sample that were processing
t= corresponds to a single data point inside a sample
X(t)=is the resulting data point,
x(t)=is the data point value before standardization,
µ(i)= is the mean value of the whole sample
σ(i)= is the standard deviation of the whole sample.
ELECTRONICS AND COMMUNICATION ENGINEERING 7
Fig2: Unfiltered data(a,c), Filterd
data(b,d)

8. Methodology (3/6)
Feature Extraction Methods:
1. Maximum peak amplitude value(PAV) - signal amplitude at the highest point,
maximum positive value
2. Maximum valley amplitude value(VAV) - signal amplitude at the lowest point,
maximum negative value
3. Maximum peak amplitude position value(PAP) – signal amplitude position value at the
highest point, maximum positive value
4. Maximum value amplitude position value(VAP) - signal amplitude position value at the
lowest point, maximum negative value.
ELECTRONICS AND COMMUNICATION ENGINEERING 8

9. Methodology (4/6)
5. Area under curve value(AUC) - summation of absolute value of the amplitude
under both positive and negative curves in both V and H channels. It can be expressed
as
x(i) = ith sample of EOG signal
N = window size for computing features.
ELECTRONICS AND COMMUNICATION ENGINEERING 9

10. Methodology (5/6)
6. Number of thresholds crossing value(TCV) - number of times that the EOG
signal passes the threshold amplitude value for both positive and negative threshold
values, in both V and H channels.
7. Number of thresholds crossing value(TCV) - measure of the signal power and
calculated as
x(i) = ith sample of EOG signal
N = window size for computing features
ELECTRONICS AND COMMUNICATION ENGINEERING 10

11. Methodology (6/6)
Classification
 Support vector machine (SVM) will be used for classification of extracting features.
 K-fold cross validation and support vector machine(SVM) will be used for splitting
and classifying the entire data respectively.
 To fit the model k-1 folds and kth fold will be used for validating the model.
 The model accuracy will be calculated as the mean of these k models.
11

12. Future work
• If the model gives better accuracy the system will be easy to connect to subsequent
communication or movement assistance systems.
• By deducting unintentional blink, noisy and flat signal, it could be possible to improve
accuracy.
ELECTRONICS AND COMMUNICATION ENGINEERING 12

13. Conclusion
• Comparison of results between different EOG signal classification.
• Best feature extraction methods are chosen.
• For classification, SVM have been chosen.
• Some work will be done to improve the accuracy of the system.
ELECTRONICS AND COMMUNICATION ENGINEERING 13

14. References(1/3)
[1]. Qi, L.J.; Alias, N. Comparison of ANN and SVM for Classification of Eye Movements in EOG Signals. J. Phys.
Conf. Ser. 2018, 971, 012012.
[2]. Guo, X.; Pei, W.; Wang, Y.; Chen, Y.; Zhang, H.; Wu, X.; Yang, X.; Chen, H.; Liu, Y.; Liu, R. A Human-Machine
Interface Based on Single Channel EOG and Patchable Sensor. Biomed. Signal. Process. Control. 2016, 30, 98–105.
[3]Kherlopian, A.; Sajda, P.; Gerrein, J.; Yue, M.; Kim, K.; Kim, J.W.; Sukumaran, M. Electrooculogram Based
System for Computer Control Using a Multiple Feature Classification Model. 4. In Proceedings of the 28th
IEEE EMBS Annual International Conference, New York, NY, USA, 30 August–3 September 2006.
[4].Wu, S.L.; Liao, L.D.; Lu, S.W.; Jiang, W.L.; Chen, S.A.; Lin, C.T. Controlling a Human–Computer Interface
System with a Novel Classification Method That Uses Electrooculography Signals. IEEE Trans. Biomed. Eng.
2013, 60, 2133–2141
[5].Heo, J.; Yoon, H.; Park, K. A Novel Wearable Forehead EOG Measurement System for Human Computer
Interfaces. Sensors 2017, 17, 1485.
[6]. Erkaymaz, H.; Ozer, M.; Orak, I.M. Detection of Directional Eye Movements Based on the Electrooculogram
˙Signals through an Artificial Neural Network. Chaos Solitons Fractals 2015, 77, 225–229.
[7].Merino, M.; Rivera, O.; Gomez, I.; Molina, A.; Dorronzoro, E. A Method of EOG Signal Processing to Detect the
Direction of Eye Movements. In Proceedings of the First International Conference on Sensor Device Technologies and
Applications, Venice, Italy, 18–25 July 2010; IEEE: Piscataway, NJ, USA, 2010; pp. 100–105.
ELECTRONICS AND COMMUNICATION ENGINEERING 14

15. References(2/3)
[8]. Aungsakul, S.; Phinyomark, A.; Phukpattaranont, P.; Limsakul, C. Evaluating Feature Extraction Methods of
Electrooculography (EOG) Signal for Human-Computer Interface. Procedia Eng. 2012, 32, 246–252.
[9]. Phukpattaranont, P.; Aungsakul, S.; Phinyomark, A.; Limsakul, C. Efficient Feature for Classification of Eye
Movements Using Electrooculography Signals. Therm. Sci. 2016, 20, 563–572.
[10]. Jayro Martínez-Cerveró ,Majid Khalili Ardali ,Andres Jaramillo-Gonzalez ,Shizhe Wu, Alessandro Tonin ,Niels
Birbaumer andUjwal Chaudhary. Open Software/Hardware Platform for Human-Computer Interface Based on
Electrooculography (EOG) Signal Classification. MDPI. Sensors 2020, 20(9), 2443
[11]. Wu, S.L.; Liao, L.D.; Lu, S.W.; Jiang, W.L.; Chen, S.A.; Lin, C.T. Controlling a Human–Computer Interface
System with a Novel Classification Method That Uses Electrooculography Signals. IEEE Trans. Biomed. Eng. 2013, 60,
2133–2141.
[12]. Rokonuzzaman, S.M.; Ferdous, S.M.; Tuhin, R.A.; Arman, S.I.; Manzar, T.; Hasan, M.N. Design of an
Autonomous Mobile Wheelchair for Disabled Using Electrooculogram (EOG) Signals. In Mechatronics; Jablonski, R.,
Brezina, T., Eds.; Springer: Berlin/Heidelberg, Germany, 2011; pp. 41–53.
[13]. Hossain, Z.; Shuvo, M.M.H.; Sarker, P. Hardware and Software Implementation of Real Time Electrooculogram
(EOG) Acquisition System to Control Computer Cursor with Eyeball Movement. In Proceedings of the 4th International
Conference on Advances in Electrical Engineering (ICAEE), Dhaka, Bangladesh, 28–30 September 2017; IEEE:
Piscataway, NJ, USA, 2017; pp. 132–137.
ELECTRONICS AND COMMUNICATION ENGINEERING 15

16. References(3/3)
[14]. Usakli, A.B.; Gurkan, S. Design of a Novel Efficient Human–Computer Interface: An Electrooculagram Based
Virtual Keyboard. IEEE Trans. Instrum. Meas. 2010, 59, 2099–2108.
[15]. Argentim, L.M.; Castro, M.C.F.; Tomaz, P.A. Human Interface for a Neuroprothesis Remotely Control.
In Proceedings of the 11th International Joint Conference on Biomedical Engineering Systems and
Technologies, Funchal, Madeira, Portugal, 19–21 January 2018; SCITEPRESS—Science and Technology
Publications: Setubal, Portugal, 2018; pp. 247–253.
ELECTRONICS AND COMMUNICATION ENGINEERING 16