Eye Tracking Using A Novel Approach

Unobtrusive and pervasive video-based eye-gaze tracking

“…Feature-based methods, on the other hand, seek to identify specific informative local features of the eye region that may be less susceptible to illumination changes and variations in viewpoint [52][53][54][55][56][57][58][59][60][61][62][63][64][65]. Detection of these local features may either exploit the grey-level differences at the feature boundaries, such as the limbus boundary [52][53][54][55], or alternatively the dark and distinctive colour of the pupil [56][57][58][59][60][61][62][63][64][65]. In order to reduce the number of eye candidates that may be captured by wide field-of-view imaging hardware and which may exhibit similar features to the eyes, feature detection is often performed within close-up eye region images [63][64][65].…”

“…Detection of these local features may either exploit the grey-level differences at the feature boundaries, such as the limbus boundary [52][53][54][55], or alternatively the dark and distinctive colour of the pupil [56][57][58][59][60][61][62][63][64][65]. In order to reduce the number of eye candidates that may be captured by wide field-of-view imaging hardware and which may exhibit similar features to the eyes, feature detection is often performed within close-up eye region images [63][64][65]. Different methods for eye detection and localisation are summarised in Figure 2 and illustrated in Figure 3.…”