الفهرس | Only 14 pages are availabe for public view |
Abstract Authentication is the process of determining whether someone or something is, in fact, who or what it is declared to be. In general, authentication methods can be broadly categorized into three groups 1) knowledge-based, which typically relies on a password. 2) object-based, which relies on a physical possession such as tokens or smart cards. 3) Identity-based, i.e. biometrics, which relies on the uniqueness of Physiological or behavioral characteristics of a person. The term biometrics has been used to denote the unique biological traits of individuals, such as face images, fingerprints, iris, voice print, etc., that can be used for identification. Since these traits cannot be stolen, lost, or forgotten, they offer better inherent security and reliability in identifying people. Recently, there is a considerable effort to replace traditional means of identification such as the use of passwords with biometric-based authentication systems. The general architecture of a traditional biometric authentication system consists of two main phases: training (enrollment) and testing (recognition). The biometric sensor captures the data from the biometric and sends biometric data to the feature extraction module that extracts selected information (features) from the data and creates a unique feature vector for the biometric sample. In the enrollment phase, these features are stored in a database as templates. In the recognition phase, the matcher receives the extracted feature vector and compares it with the feature vector of all templates which are already stored in the database (for identification) or with one specific template (for verification). iii Biometrics authentication systems are gaining wide-spread popularity in recent years due to the advances in sensor technologies as well as improvements in the matching algorithms that make the systems both secure and cost-effective. A biometric system can operate in verification and identification modes. In computer science, in particular, biometrics is used as a form of identity access management and access control. It is also used to identify individuals in groups that are under surveillance. Face recognition is the one of the most important topic in biometrics authentication and it has drawn attention of the research community. In addition, it has become one of the most active applications of visual pattern recognition due to its potential value for law enforcement, surveillance, and human-computer interaction. Nowadays, various systems are able to properly recognize people based on their face image. The recognition of a face in a single image involves at least these three stages: Face detection, Feature extraction, and Face recognition. Face recognition systems solves, mainly, two kinds of problems [21]: identification problems (answer the question: “Who am I?”), where the input to the system is unknown face, and the system gives back the determined identity from a database of known individuals; and verification problems (answer the question: “Am I who I say I am?”), which verify that the individual is who he claims to be. Several surveys papers [12, 22, 23, 24, 25, 26] and books [27, 28, 29] on human and machine recognition of faces have been published, which gave very good reviews on face recognition. In order to find out the true invariant for face recognition, researchers have developed recognition algorithms such as Principal Component Analysis (PCA, also known as iv Eigenfaces) [18, 20], Fisher Discriminant Analysis (FDA, also known as Fisherfaces, Linear Discriminant Analysis) [15], Self Organizing Map and Convolutional Network (SOM+CN) [30], template matching [31], Modular PCA [32], Line Edge Maps (LEM) [16], Elastic Bunch Graph Matching (EBGM) [33], Directional Corner Point (DCP) [34], Local Binary Patterns (LBP) [35], and etc. Face recognition is a challenging task because of variable factors like alterations in scale, location, pose, facial expression, occlusion, lighting conditions and overall appearance of the face. With the synergy of efforts from researchers in diverse fields different frameworks have evolved for solving the problem of face recognition. In this work, we develop some techniques to answer the question ”Who am I? ”. For this purpose, we propose a face recognition technique based on edge detection and distance similarity measures. Accordingly, mean square error and correlation coefficient are used as a distance similarity measures for matching. A second technique for face recognition based on genetic programming is proposed. An advantage of the proposed techniques is that they are not affected by face recognition aspects such as lighting condition, varying facial expression, and varying pose. In addition, the results demonstrate that the proposed techniques can obtain better performances than other existing face recognition techniques. The thesis comprises five chapters, these are organized as follows Chapter one: gives an overview of biometrics, biometrics types, and their applications. Chapter two: surveys face recognition algorithms and face recognition challenges. Chapter three: reviews edge detection and proposes face recognition technique based on edge detection and distance similarity measures. Also, comparisons between this proposed technique and other existing techniques are introduced. Chapter four: reviews genetic programming and proposes a face recognition technique based on genetic programming. Also, comparisons between this proposed technique and other existing techniques are considered. Chapter five: draws the conclusion and suggests headlines for future work. |