Eyeblink application released

Please follow www.blinkingmatters.com

portrait








Objective of research project

Purpose of this project is to help people with dry eye sondrome and at the same time connect research in computer vision and medicine. We cooperate with Ofhtalmology clinic of Faculty of Medicine. The aim of the project is to develop fast algorithms, which could be eventually used in application, which could be used to help people to prevent and help with dry eye syndrome.

Dry eye syndrome

Up to 70% of people who work in front of computer screen every day suffer felt at least one of the symptoms of dry eye like dryness, redness and burning or sandy-gritty eye irritation, sensitivity to light and eye fatigue. Eye drops deals with these symptoms, but they can never prevent them. Sufficient eye blink frequency is the only way to prevent them. A microscopic leayer of tear film protects and moisturizes the eye surface. With each blink the tear film is renewed. A healthy human blinks 10 to 15 times per minute. Because of monitor, computer user blinks tremendously less (2 to 7 times) due to which dry eye syndrome often occurs.

Computer vision algorithms and Webcam!

We are working on a program which will measure the eye blink frequency. Our aim is to use a webcam (commonly accessible hardware) and computer vision algorithms. Webcam is pointed to the user's face to see his eyes. Our program will detect individual eye blinks and calculate the eye blink frequency. If the frequency will be lower than normal healthy human, program will notify the user in a noticeable but non-intrusive way (for example small image of blinking eye on the right side of the screen). This notification will be dismissed by the following eye blink of the user, which would not disturb his work. In this way we want to learn the user to blink consciously and regularly to prevent and heal (in a case of already having eye health issues) the computer vision syndrome. Graphical user interface is very important component; the purpose is to bring intuitiveness and comfort. Program should not slow down the work of the user and that is why the notification to remind to blink consciously should not disturb him. Eventual blink detection will dismiss the notification and the user does not have to move mouse and click or press any key.


Project support

I received E-talent grant from Tatra bank foundation to cover the cost on hardware during development process. I also win an Innovative voucher 2013 organized by SLOVAK CENTRE OF SCIENTIFIC AND TECHNICAL INFORMATION. I also win Social Impact Award with this project.

People

Faculty of Informatics and Information Technology, Slovak University of Technology

Ing. Andrej Fogelton
Milos Palo
Bc. Tomas Drutarovsky [not active anymore]
Peter Sarissky [not active anymore]
Marek Hlava [not active anymore]
Bc. Patrik Polatsek [not active anymore]

Ofhtalmology Clinic, Faculty of Medicine, Comenius University

doc. Mudr. Alena Furdova PhD.
Mudr. Jan Juhas
Mudr. Ivana Krcova [not active anymore]
Mudr. Kristina Horkovicova [not active anymore]
Veronika Farkasova [not active anymore]

Project output

Thesis

Bc. Patrik Polatsek, BLINK RATE TRACKING OF COMPUTER USER, [deans's award for the best bachelor thesis]
Bc. Tomas Drutarovsky, Driver drowsiness detection, [dean's letter of praise for excellent thesis]
Bc. Peter Sarissky

Papers

Polatsek, P.: Eye Blink Detection, Proceedings of 9th Student Research Conference in Informatics and Information Technologies, Bratislava, Slovakia, April 23, 2013, Publisher STU

Drutarovsky T., Fogelton A.: Eye Blink Detection using Variance of Motion Vectors, Proceedings of 2th workshop on Assistive Computer Vision and Robotics in ECCV 2014, Springer

Software

Video recording and pacient database software at ophtalmology clinic
Portable video recording software
we tried 8 eyeblink detection algorithms
application prototype [in progress]