Face Detection and Pose Estimation Based on Evaluating Facial Feature Selection

Abstract

The detection of faces is one of the most requesting fields of research in image processing and Visual estimation of head pose is desirable for computer vision applications such as face recognition, human computer interaction, and affective computing. In this paper, we propose completed method for face pose estimation, face and face parts detection, feature extraction, tracking. This paper proposes using an improved AdaBoost algorithm, which is much better than normal AdaBoost. We use the de-facto Viola-Jones method for face and face part detection. From the robustness property of Haar-like feature, we first construct the strong classifier more effective to detect rotated face, and then we propose a novel method that can reduce the training time. We adopt affine motion model estimation as a tracking method. The combination enables efficient detection around the search area limited by tracking. Experimental results demonstrated its effectiveness and robustness against different types of detection and pose estimation in the input face images, including faces that appear in a wide range of image positions and scales, and also complex backgrounds, occlusions, illumination variations and multi-pose head images.