Rapid Eye Detection Method for Non-Glasses Type 3D Display on Portable Devices

Abstract

Eye detection is a critical task in many applications, especially for portable single user 3D display systems of the non-glasses type. However, existing methods either have high computational complexity or a low detection rate. In addition, these methods do not work with those using eye glasses for portable devices. In this paper, we propose a rapid eye detection method for portable devices. We propose a novel operator that measures contrast features to compute the difference in intensity between the iris region and its neighboring regions. The probability of an identified area being the eye position is calculated by our proposed contrast operator and the a priori geometric information of the two eyes. Experimental results show that the proposed method successfully detects the eyes in facial images regardless of illumination changes or wearing glasses. The average accuracy of the proposed method is 92.7%, which is higher than conventional methods. In addition, the proposed method can be run at an average rate of 12 fps on a portable device.