A Method for Accurate Localization of Facial Features

Facial features localization, as an important technique in human face analyzing, is specialized in searching for facial features (eyes, nose, mouth, etc) within a given region in an image. Although there are many methods to locate the facial features, most of them are computationally expensive. This paper presents a simple method that uses spatial filter for accurate feature detection in face images. When using this algorithm in complex images (images that contain both human faces and some other objects), we find that, there are many false positive candidates. In order to eliminate these candidates, we use a new variant of AdaBoost classifiers based face detection to detect the face regions, thus limit the regions that feature localization algorithm analyzed.