EYE MOVEMENT AND BLINK DETECTION FOR SELECTING MENU ON-SCREEN DISPLAY USING PROBABILITY ANALYSIS BASED ON FACIAL LANDMARK

摘要

The menu selection on a monitor screen is usually achieved using a remote control, mobile device, mouse, touch system embedded into a monitor screen, or a keypad or keyboard. Users unable to move their limbs can still utilize their eyes as an alternative method for selecting an on-screen menu display. As a breakthrough for people with movement disabilities, computer vision allowing the detection of eye movements can be implemented as a menu selection tool of a display monitor. The detection of the eye position consists of the left, right, and middle areas. Meanwhile, as an additional necessity for the execution process, eye blinking also needs to be detected by the system. The method proposed in this study for the detection of the eye movement position during the initial step includes the segmentation process and a calculation of the probability of a white pixel analysis based on the facial landmarks. Eye blinking can be detected by calculating the value between the horizontal and vertical lines in the eye area. The proposed method for detecting eye movements in three areas (left, center, and right) achieves an average accuracy of 88.1 %. Furthermore, the average accuracy for detecting eye blinking is 90.5%.