Application of head flexion detection for enhancing eye gaze direction classification

Annu Int Conf IEEE Eng Med Biol Soc. 2014:2014:966-9. doi: 10.1109/EMBC.2014.6943753.

Abstract

Extensive research has been conducted on the tracking and detection of the eye gaze and head movement detection as these aspects of technology can be applied as alternative approaches for various interfacing devices. This paper proposes enhancements to the classification of the eye gaze direction. Viola Jones face detector is applied to first declare the region of the eye. Circular Hough Transform is then used to detect the iris location. Support Vector Machine (SVM) is applied to classify the eye gaze direction. Accuracy of the system is enhanced by calculating the flexion angle of the head through the utilization of a microcontroller and flex sensors. In case of rotated face images, the face can be rotated back to zero degrees through the flexion angle calculation. This is while Viola Jones face detector is limited to face images with very little or no rotation angle. Accuracy is initiated by enhancing the effectiveness of the system in the overall procedure of classifying the direction of the eye gaze. Therefore, the head direction is a main determinant in enhancing the control method. Different control signals are enhanced by the eye gaze direction classification and the head direction detection.

MeSH terms

  • Algorithms*
  • Biomechanical Phenomena
  • Female
  • Fixation, Ocular / physiology*
  • Head Movements / physiology*
  • Humans
  • Male