Sensor technology is getting advanced day-by-day which aids for the use of biometric technology in almost every area and wide variety of types of biometric systems are being used to achieve various functionalities. Human face is one of the highly preferred means to authenticate an individual. Face recognition is sensitive to aspects like pose variation, illumination and occlusion. In this paper, the authors propose a fully automatic face recognition system which is robust to occlusions. They basically consider two problems: occlusion handling for surface registration and recognition with missing data.