BEST LOW-COST METHOD FOR REAL TIME DETECTION OF THE EYE AND GAZE TRACKING

Abstract

One important field of computer vision research is the study of gaze tracking. It emphasizes practical usage and the interaction between people and technology. The demand for inexpensive techniques has increased recently due to new eye tracking applications. An essential component in tracking the gaze's direction is the ocular area. This research proposes a number of novel eye-tracking techniques that use techniques to identify the eye region and the direction of gaze. Eye tracking may be done using unmodified cameras without the requirement for specific hardware or software. The eye area was identified using either the Haar cascade approach or facial landmarks. Furthermore, the orientation of the eye was ascertained using the engineering technique, which relies on distances identifying the iris area, and the direct method, which is based on the convolutional neural network model. Two engineering approaches are used in the paper: the iris region is divided into five parts, with the blackest region indicating the look direction, and the gaze direction junction point is identified by drawing perpendicular lines on the iris region. While engineering techniques increase their efficacy in broad mobility, the suggested network model has demonstrated efficacy in predicting the eye's gaze direction in restricted mobility.