Robust feature extraction for facial image quality assessment

Abstract

With the aim of developing an automatic system to verify if facial images meet ISO/IEC 19794-5 standards and ICAO requirements, this paper proposes a robust method to detect facial features including two eye centers and four lip features. The proposed method restricts the areas where facial features are observed by using the skin color and shape characteristic of faces. Two eye centers are detected independently in the restricted area by means of the circular filters. The use of circular filters makes the algorithm robust to head poses and occlusions, which are the main factors of unsuccessful eye detections. To accurately detect lip features regardless facial expressions and the presence of beard or mustache, the proposed method fuses edge and color information together. An experiment was performed on a subset of the FERET database, and the experimental results demonstrated the accuracy and robustness of the proposed method. © 2011 Springer-Verlag.