Real-Time Video-Based Eye Blink Detection

Eye is an important part of the human body and, it plays a fundamental role in vision based human-computer (HCI) interaction technology. Eye blinking is a major feature in HCI and, its application in such systems has been increasingly grown. This paper presents an adaptive blink detection algorithm. The advantages of this algorithm are simplicity, accuracy, very low computational cost and robustness against lighting conditions. This algorithm is based on very simple image processing techniques. The first step of blink detection method is eye detection. To accomplish this, we suppose that a fairly large face image is available. Each frame of the input video is processed and the location of eye is found. The next step calculates a value to determine the state of eye and then by analyzing it, eye blink is detected. The accuracy of our method is tested by providing a data set of several people and comparing the results to some existing methods. The experimental results show that the proposed method has 94.45% detection accuracy with overall accuracy of 99.84%. The average blink detection time for a sample is less than 0.1ms which is suitable for real-time applications.