FACE RECOGNITION: HOLISTIC APPROACHES AN ANALYTICAL SURVEY

Abstract: Face recognition has been in spot light for last few decades by keeping in view its increasing usage in real world applications, still challenges are there to meet, especially in real world applications. A lot of work has been reported on face recognition during the recent decades, some of which have also come up with their modifications. This paper presents a detailed analysis on the importance and application of face recognition technology, mentioning the factors affecting its applicability in real life. In addition, several face recognition techniques along with their experimental results are also discussed. The issues that still need to be addressed are mentioned here as well. Key words: Face, Recognition, Biometrics, Survey, Holistic INTRODUCTION Face recognition technology[1] has been considered seriously by the researchers in the past few years keeping in view its escalating usage in law enforcement and commercial applications[2; 3]. This technology is the result of a long research over a period of 30 years. Although currently available face recognition systems have become quite mature; still their accomplishment is restricted by the problems and situations of real life environment, for there are many factors that affect face recognition [4,5]. For instance, problems of pose and illumination [6] are still a big challenge. Because of this Biometrics, pattern recognition and computer vision communities are paying special attention towards research work going on in this particular field, as a result of which face recognition and image processing have become prominent research areas in the global market. Different methods have been developed based on these techniques [7], some of which are subspace analysis, Log polar gabor, elastic graph matching, neural net-work, support vector machine(SVM), deformable intensity surface, morph able model, etc. Face recognition process requires that an individual is identified successfully. This technique is mainly used in applications providing security and surveillance. Researchers have developed different techniques and algorithms which are quite efficient in the process of recognizing faces under constraints like lighting and pose problems. One approach for handling pose variations and 3S problems for face recognition can be analyzed in [8].There are number of application for face recognition systems. The very first use of such systems is in the security management systems for criminal identification. Face recognition systems can be used in combination with surveillance cameras in order to increase the security system. Pattern recognition is another important application of these systems. Face recognition systems can be used in diverse vicinities of science for evaluating an entity with a set of entities. Face recognition on the bases of general view point with different backgrounds, illumination changes, different facial expressions and handling age factor is one of the biggest challenges of such systems[9].Together with challenges face recognitions systems are facing some criticisms which include weaknesses, privacy issues and effectiveness. Weaknesses contain many situations including pose, aging and illumination factors. In such situations the available methods are not much effective and efficient. The criticism in terms of privacy issues includes the compromise of privacy through the use of surveillance cameras. Effectiveness is criticized on the bases of inefficiency of such systems to identify a criminal.