Automated Quantification of Eye Blink Rate Using VIOLA–JONES Algorithm

Automated Quantification of Eye Blink Rate Using VIOLA–JONES Algorithm

Mohammad Hamdan (Department of Computer Science, Yarmouk University, Irbid, Jordan and MACS, Heriot-Watt University, Dubai, UAE) and Hisham A. Shehadeh (Department of Computer System and Technology, University of Malaya, Kuala Lumpur, Malaysia)
Copyright: © 2018 |Pages: 14
DOI: 10.4018/IJTD.2018100102

Abstract

In this article, we have proposed a novel tool that helps to objectively quantify eye blink rate. Using the proposed algorithm, a threshold for normal blink rate can be set to test those who have to reduce eye blink rate and are prone to ocular surface dryness. The statistical results show excellent agreement between software-detected number of blinks and visually measured with 90% accuracy for the participants. In addition, the comparison between our tool and other approaches of eye blink monitoring shows that our tool is competitive with only 5% wasted blinks.
Article Preview
Top

Introduction

The ocular surface of the eye comprises of the cornea, the conjunctiva and the tear film. The tear film is considered the most anterior structure of the eye that normally covers the cornea and the conjunctiva, which prevents ocular surface dryness. Tear film distribution on the ocular surface is mainly managed by eye blinks which includes spontaneous and complete closure and the opening of the eyelids (Doane, 1980).

Eye blink is a normal ocular response that is important for tear film distribution on the ocular surface and it is a defence mechanism protecting the eye from the invasion of microbes and foreign body (Doane, 1980).

The normal blink rate ranges from 10-15 times per minute and usually varies among normal people in different cognitive processes, such as rest, conversation and doing a near activity (Bentivoglio et al., 1970). Changes in the normal blink rate either in a decrease or increase or incomplete closure may affect the tear film or lead to ocular symptoms such as discomfort, pain, and dryness (Tsubota and Nakamori, 1993). Figure 1 shows symptoms of the dry eye like dilated blood vessels near the surface of the eye.

Decreased blink rate may occur during reading, computer usage and as a result of disease (Portello, Rosenfield, & Chu, 2013; Doughty, Naase, & Button, 2009; Orchard & Stern, 1991). While increased rate of blinking may occur in several nervous system diseases, such as stroke, Parkinson disease and dyskinesia (Reddy et al., 2013).

For other purposes, assessment of blink parameters including blink rate and completeness has been suggested to take into consideration in the design of Video Display Terminal (VDT) (Tsubota et al., 2017) to increase the user’s comfort-ability (Portello, Rosenfield, & Chu, 2013).

Figure 1.

Dry eye - dilated blood vessels near the surface of the eye

IJTD.2018100102.f01

According to the literature, eye blink rate can be counted incidentally using several methods. First, using eye-movement trackers where eye blink is monitored when the position of the eye is hidden during the blinking event (Zaman & Doughty, 1997). The second way is with Electro-Encephalo-Gram (EEG), which are normally detected and discounted (Siswandari & Xiong, 2015). For accurate detection of eye movement, special electrodes are often applied in EEG work, which can incidentally record blinks (Siswandari & Xiong, 2015). The third method of blinks quantification is detecting eye blink through contact lens sensors (Gisler et. al., 2015). The fourth method includes video captures from subjects and subsequent actual-time analysis of the video. When the cornea completely disappears then this counts as a complete blink (Gisler et al., 2015).

From the previous discussion, we can see that various researchers look at eye blink rate for different purposes. Next, we highlight the most important cases. Some of them concentrate on proposing a new algorithm for detecting eye blink rate. Others look at using different devices to analyse electroencephalographic (EEG) signals in order to determine the status of eyes; (blink or not blink). Furthermore, some of them use different types of sensors such as blink sensors. However, these studies lack determining the abnormality status of eyes that lead to vision diseases such as dry eye. For this reason, in this paper, we propose a tool to detect, analyze and quantify eye blink rate. We then compare the detected number of blinks with a threshold value to determine the person’s eyes status. In addition, users can use this tool to give them an alarm of eye problem in case their blink rate is less than normal rate. The threshold value and the tools used will be discussed in the following sections.

Complete Article List

Search this Journal:
Reset
Open Access Articles
Volume 12: 4 Issues (2021): Forthcoming, Available for Pre-Order
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing