Eye State Detection and Eye Sequence Classification for Paralyzed Patient Interaction

New approaches of eye state detection and eye sequence identification for computer interface of paralyzed patients are proposed. In this work, patients can interact via sequences of four eye states that are close, forward-glance, rightward-glance, and leftward-glance states. To detect the eye states, eye images are firstly segmented by using FCM clustering scheme in a feature space of RGB color components and pixel coordinate. Features are extracted from image projection and bottom edge curve of the segmented eye image. Then, the eye state is recognized by using SVM. The eye state sequences can be identified by using modified Levenshtein distances between unknown eye sequences and prototypes of command sequences which are generated using HMM. The experiments show that accuracies of eye state classification are 95.37% for four-class classification and 99.47% for open-close state classification. An accuracy of the sequence pattern recognition is 91.32% which can be concluded that the proposed method works effectively for the purpose of paralyzed patient interaction.