Robust head pose estimation using locations of facial components

Head pose estimation is used in many applications such as driver fatigue estimation. Further, it is considered as the primary step in face recognition systems. In this work, we propose a novel algorithm that estimates head pose by considering the locations of facial components such as eyes and mouth. At first, we use two algorithms for developing a new image in which the eyes and mouth are emphasized in face image. Next, in order to extract proper features, we consider the sum of pixels in each column of new image. To classify the extracted features, we apply k-nearest neighborhood (k-nn) and support vector machine (SVM) classifiers separately. We use FERET face database to evaluate the performance of the proposed algorithm. Experimental results demonstrate the efficiency of the proposed method.