Comprehensive Study of Face Recognition Using Feature Extraction and Fusion Face Technique

Introduction

Face recognition is a process by which the identity of a person is determined from the face images stored in a face database (Dey et al., 2014). Face recognition is one of the most successful applications of image analysis. In present scenario, face recognition plays a major role in commercial and law enforcement applications, such as, surveillance system, passport, security, personal information accesses, human machine interaction, etc. (Yuille et al., 1989; Zitová et al., 2003). At present, very reliable methods of biometric personal identification exist. The fingerprint analysis, retinal or iris scans, etc., are examples of the biometric personal identification method. However, these methods rely on the active cooperation of the participants (Graham et al., 1998). The person identification system, which is based on analysis of frontal or profile images of the face, is very effective without the participant’s cooperation or knowledge. Face perception is an important part of the capability of human perception system (Alvarado et al., 2006). It is also a routine task for human. It is a true challenge to build a computer system which parallels human ability to recognize faces. Therefore, face recognition has become an active research area, and it attracts researchers from the field of image processing, pattern recognition, neural networks, computer vision, etc. (Zhou et al., 2014). Although, presently the face recognition system has reached a certain level of maturity, but the success of the face recognition system is limited by the conditions imposed by many real-world applications. For example, face recognition in an outdoor environment with variations in illumination, and/or pose remains a very challenging problem (Fraser et al., 1998; Sing 2015). Therefore, the present systems are still far away from the capability of the human perception system. The face recognition system can be developed as a three-step process (Li, 2014). The first step of the face recognition system is face detection (Keller et al., 1985). Face detection is the process of extracting face region from the input scene (Yang et al., 2004). It has many applications in face tracking (Zhuang & Dai, 2007), pose estimation, compression, human-computer-interaction (HCI) system, etc. The next step of the face recognition system is feature extraction (Tan et al., 2006), which acquires relevant facial features from the face images (Adini et al., 1997; Zou et al., 2007). Features are properties which describe the whole face image (Xu et al., 2013). Feature extraction process must be efficient enough in terms of computing time and memory usage (Kwak et al., 2005). There are many applications of the feature extraction process in facial feature tracking, emotion recognition, gaze estimation, and human-computer-interaction (HCI) system (Nandakumar 2008; Cament et al., 2015; Shen et al., 2004). The face detection and feature extraction are often performed simultaneously (Er et al., 2002; Bartlet et al., 2002). The final step is face recognition. In this phase, the face images are identified or verified by applying the extracted facial features on some classifiers (Zhao et al., 2012).