Eye Shape and Corners Detection in Periocular Images Using Particle Filters

Abstract

The eyes are the most preeminent features of the human face and the ability to accurate the eye landmarks is crucial to a variety of application domains. In this paper, we present a probabilistic method to detect the eye shape in periocular images based on particle filters. The proposed method does not need any prior information about the position of the iris and there is no need for initialization. The eyes are modeled by a simple feature vector that generates two parabolas for the upper and lower eyelid. In order to ensure the robustness of the solution, several measurement cues are fused together when computing the score of a hypothetical eye shape. The proposed method was extensively evaluated on a publicly available database.