Tight bounded localization of facial features with color and rotational independence

Human face detection plays an important role in applications such as video surveillance, human computer interface, face recognition and face image database management. In this paper, we propose a novel facial feature detection algorithm for various face image types, conditions, invariant rotation, and any appearances. There are three main steps. First, Radon transform is used for face angle detection on rotated image. Subsequently, the feature regions are detected using Neural Visual Model (NVM). Finally, using image dilation and Radon transform, the facial features are extracted from the detected regions. Input parameters are obtained from the face characteristics and the positions of facial features not including any intensity informations. Our algorithm is successfully tested with various types of faces which are color images, gray images, binary images, wearing the sunglasses, wearing the scarf, lighting effect, low-quality images, color and sketch images from animated cartoon, rotated face images, and rendered face images.