Facial feature detection: A facial symmetry approach

Abstract

Nowadays face detection plays an important role in recognition, emotion recognition, computer-human interaction, etc. This paper presents a novel method for the detection of facial features in images. The main objective is to develop a fully automatic facial feature detection system. The method proposed in this paper uses a combination of methods to detect facial features. It first uses the Viola-Jones methods to identify possible regions of interest subsequently use calculations based on the symmetric property of the human face to detect the true facial features. A comparison between the Viola-Jones algorithm and the proposed algorithm has been performed and it shows that our method in combination with Viola-Jones increases the accuracy of detection considerably.