A new approach for rotated face detection

Abstract

Human face detection has always been an important problem for face recognition and face tracking. Though considerable attempts have been made to detect and localize faces, these approaches have made assumptions that restrict their extension to more general cases. In this paper we design a novel method to detect a rotated face on the basis of image edge information. Considering the efficiency problem, we propose two key techniques in our approach: first, three points based RHT is applied to detect an ellipse; moreover, we speedup the template matching procedure through calculating the orientation histogram distribution and locating the symmetry axis of a rotated face. In the end, we will validate our idea by several experiment results.