Dissertations / Theses on the topic 'Iris pattern'

This research focuses on iris recognition, the most accurate form of biometric identification. The robustness of iris recognition comes from the unique characteristics of the human, and the permanency of the iris texture as it is stable over human life, and the environmental effects cannot easily alter its shape. In most iris recognition systems, ideal image acquisition conditions are assumed. These conditions include a near infrared (NIR) light source to reveal the clear iris texture as well as look and stare constraints and close distance from the capturing device. However, the recognition accuracy of the-state-of-the-art systems decreases significantly when these constraints are relaxed. Recent advances have proposed different methods to process iris images captured in unconstrained environments. While these methods improve the accuracy of the original iris recognition system, they still have segmentation and feature selection problems, which results in high FRR (False Rejection Rate) and FAR (False Acceptance Rate) or in recognition failure. In the first part of this thesis, a novel segmentation algorithm for detecting the limbus and pupillary boundaries of human iris images with a quality assessment process is proposed. The algorithm first searches over the HSV colour space to detect the local maxima sclera region as it is the most easily distinguishable part of the human eye. The parameters from this stage are then used for eye area detection, upper/lower eyelid isolation and for rotation angle correction. The second step is the iris image quality assessment process, as the iris images captured under unconstrained conditions have heterogeneous characteristics. In addition, the probability of getting a mis-segmented sclera portion around the outer ring of the iris is very high, especially in the presence of reflection caused by a visible wavelength light source. Therefore, quality assessment procedures are applied for the classification of images from the first step into seven different categories based on the average of their RGB colour intensity. An appropriate filter is applied based on the detected quality. In the third step, a binarization process is applied to the detected eye portion from the first step for detecting the iris outer ring based on a threshold value defined on the basis of image quality from the second step. Finally, for the pupil area segmentation, the method searches over the HSV colour space for local minima pixels, as the pupil contains the darkest pixels in the human eye. In the second part, a novel discriminating feature extraction and selection based on the Curvelet transform are introduced. Most of the state-of-the-art iris recognition systems use the textural features extracted from the iris images. While these fine tiny features are very robust when extracted from high resolution clear images captured at very close distances, they show major weaknesses when extracted from degraded images captured over long distances. The use of the Curvelet transform to extract 2D geometrical features (curves and edges) from the degraded iris images addresses the weakness of 1D texture features extracted by the classical methods based on textural analysis wavelet transform. Our experiments show significant improvements in the segmentation and recognition accuracy when compared to the-state-of-the-art results.

Identication of cartridge is very essential in the field of forensics, military or people who collect ammunitions. The cartridges can beidentied by their headstamps.This thesis presents work on identification and matching of cartridge headstamp from the image. The Libor Masek's open source iris recognition algorithm is considered for the identification of cartridge pattern from the image.The dataset is devoleped with the cartridge headstamp patterns and matching of cartridge headstamp patterns is implemented. For matching of the cartridge pattern the Hamming distance is considered as the metric to differentiate interclass and intraclass comparisons. Variance is used as a criteria to discard the unwanted areas of the cartridge headstamp pattern.Four distinct cartridge headstamp patterns are considered. Three cartridges of each headstamp pattern are considered for intra class comparisons. The validation of the method is performed.

Nature has created complex biological structures that exhibit intelligent behaviour through an evolutionary process. Thus, intelligence and evolution are intimately connected. This has inspired evolutionary computation (EC) that simulates the evolutionary process to develop powerful techniques such as genetic algorithms (GAs), genetic programming (GP), evolutionary strategies (ES) and evolutionary programming (EP) to solve real-world problems in learning, control, optimization and classification. GP discovers the relationship among data and expresses it as a LISP-S expression i.e., a computer program. Thus the goal of program discovery as a solution for a problem is addressed by GP in the framework of evolutionary computation. In this thesis, we address for the first time the problem of applying GP to mu1ticategory pattern classification. In supervised pattern classification, an input vector of m dimensions is mapped onto one of the n classes. It has a number of application areas such as remote sensing, medical diagnosis etc., A supervised classifier is developed by using a training set that contains representative samples of various classes present in the application. Supervised classification has been done earlier with maximum likelihood classifier: neural networks and fuzzy logic. The major considerations in applying GP to pattern classification are listed below: (i) GP-based techniques are data distribution-free i.e., no a priori knowledge is needed abut the statistical distribution of the data or no assumption such as normal distribution for data needs to be made as in MLC. (ii) GP can directly operate on the data in its original form. (iii) GP can detect the underlying but unknown relationship that mists among data and express it as a mathematical LISP S-expression. The generated LISP S-expressions can be directly used in the application environment. (iv) GP can either discover the most important discriminating features of a class during evolution or it requires minor post-processing of the LISP-S expression to discover the discriminant features. In a neural network, the knowledge learned by the neural network about the data distributions is embedded in the interconnection weights and it requires considerable amount of post-processing of the weights to understand the decision of the neural network. In 2-category pattern classification, a single GP expression is evolved as a discriminant function. The output of the GP expression can be +l for samples of one class and -1 for samples of the other class. When the GP paradigm is applied to an n-class problem, the following questions arise: Ql. As a typical GP expression returns a value (+l or -1) for a 2-class problem, how does one apply GP for the n-class pattern classification problem? Q2. What should be the fitness function during evolution of the GP expressions? Q3. How does the choice of a function set affect the performance of GP-based classification? Q4. How should training sets be created for evaluating fitness during the evolution of GP classifier expressions? Q5. How does one improve learning of the underlying data distributions in a GP framework? Q6. How should conflict resolution be handled before assigning a class to the input feature vector? Q7. How does GP compare with other classifiers for an n-class pattern classification problem? The research described here seeks to answer these questions. We show that GP can be applied to an n-category pattern classification problem by considering it as n 2-class problems. The suitability of this approach is demonstrated by considering a real-world problem based on remotely sensed satellite images and Fisher's Iris data set. In a 2-class problem, simple thresholding is sufficient for a discriminant function to divide the feature space into two regions. This means that one genetic programming classifier expression (GPCE) is sufficient to say whether or not the given input feature vector belongs to that class; i.e., the GP expression returns a value (+1 or -1). As the n-class problem is formulated as n 2-class problems, n GPCEs are evolved. Hence, n GPCE specific training sets are needed to evolve these n GPCEs. For the sake of illustration, consider a 5-class pat tern classification problem. Let n, be the number of samples that belong to class j, and N, be the number of samples that do not belong to class j, (j = 1,..., 5). Thus, N1=n2+n3+n4+n5 N2=n1+n3+n4+n5 N3=n1+n2+n4+n5 N4=n1+n2+n3+n5 N5=n1+n2+n3+n4 Thus, When the five class problem is formulated as five 2-class problems. we need five GPCEs as discriminant functions to resolve between n1 and N1, n2 and N2, n3 and N3, n4 and N4 and lastly n5 and N5. Each of these five 2-class problems is handled as a separate 2-class problem with simple thresholding. Thus, GPCE# l resolves between samples of class# l and the remaining n - 1 classes. A training set is needed to evaluate the fitness of GPCE during its evolution. If we directly create the training set, it leads to skewness (as n1 < N1). To overcome the skewness, an interleaved data format is proposed for the training set of a GPCE. For example, in the training set of GPCE# l, samples of class# l are placed alternately between samples of the remaining n - 1 classes. Thus, the interleaved data format is an artifact to create a balanced training set. Conventionally, all the samples of a training set are fed to evaluate the fitness of every member of the population in each generation. We call this "global" learning 3s GP tries to learn the entire training set at every stage of the evolution. We have introduced incremental learning to simplify the task of learning for the GP paradigm. A subset of the training set is fed and the size of the subset is gradually increased over time to cover the entire training data. The basic motivation for incremental learning is to improve learning during evolution as it is easier to learn a smaller task and then to progress from a smaller task to a bigger task. Experimental results are presented to show that the interleaved data format and incremental learning improve the performance of the GP classifier. We also show that the GPCEs evolved with an arithmetic function set are able to track variation in the input better than GPCEs evolved with function sets containing logical and nonlinear elements. Hence, we have used arithmetic function set, incremental learning, and interleaved data format to evolve GPCEs in our simulations. AS each GPCE is trained to recognize samples belonging to its own class and reject samples belonging to other classes a strength of association measure is associated with each GPCE to indicate the degree to which it can recognize samples belonging to its own class. The strength of association measures are used for assigning a class to an input feature vector. To reduce misclassification of samples, we also show how heuristic rules can be generated in the GP framework unlike in either MLC or the neural network classifier. We have also studied the scalability and generalizing ability of the GP classifier by varying the number of classes. We also analyse the performance of the GP classifier by considering the well-known Iris data set. We compare the performance of classification rules generated from the GP classifier with those generated from neural network classifier, (24.5 method and fuzzy classifier for the Iris data set. We show that the performance of GP is comparable to other classifiers for the Iris data set. We notice that the classification rules can be generated with very little post-processing and they are very similar to the rules generated from the neural network and C4.5 for the Iris data set. Incremental learning influences the number of generations available for GP to learn the data distribution of classes whose d is -1 in the interleaved data format. This is because the samples belonging to the true class (desired output d is +1) are alternately placed between samples belonging to other classes i.e., they are repeated to balance the training set in the interleaved data format. For example, in the evolution of GPCE for class# l, the fitness function can be fed initially with samples of class#:! and subsequently with the samples of class#3, class#4 and class#. So in the evaluation of the fitness function, the samples of class#kt5 will not be present when the samples of class#2 are present in the initial stages. However, in the later stages of evolution, when samples of class#5 are fed, the fitness function will utilize the samples of both class#2 and class#5. As learning in evolutionary computation is guided by the evaluation of the fitness function, GPCE# l gets lesser number of generations to learn how to reject data of class#5 as compared to the data of class#2. This is because the termination criterion (i.e., the maximum number of generations) is defined a priori. It is clear that there are (n-l)! Ways of ordering the samples of classes whose d is -1 in the interleaved data format. Hence a heuristic is presented to determine a possible order to feed data of different classes for the GPCEs evolved with incremental learning and interleaved data format. The heuristic computes an overlap index for each class based on its spatial spread and distribution of data in the region of overlap with respect to other classes in each feature. The heuristic determines the order in which classes whose desired output d is –1 should be placed in each GPCE-specific training set for the interleaved data format. This ensures that GP gets more number of generations to learn about the data distribution of a class with higher overlap index than a class with lower overlap index. The ability of the GP classifier to learn the data distributions depends upon the number of classes and the spatial spread of data. As the number of classes increases, the GP classifier finds it difficult to resolve between classes. So there is a need to partition the feature space and identify subspaces with reduced number of classes. The basic objective is to divide the feature space into subspaces and hence the data set that contains representative samples of n classes into subdata sets corresponding to the subspaces of the feature space, so that some of the subdata sets/spaces can have data belonging to only p classes (p < n). The GP classifier is then evolved independently for the subdata sets/spaces of the feature space. This results in localized learning as the GP classifier has to learn the data distribution in only a subspace of the feature space rather than in the entire feature space. By integrating the GP classifier with feature space partitioning (FSP), we improve classification accuracy due to localized learning. Although serial computers have increased steadily in their performance, the quest for parallel implementation of a given task has continued to be of interest in any computationally intensive task since parallel implementation leads to a faster execution than a serial implementation As fitness evaluation, selection strategy and population structures are used to evolve a solution in GP, there is scope for a parallel implementation of GP classifier. We have studied distributed GP and massively parallel GP for our approach to GP-based multicategory pattern classification. We present experimental results for distributed GP with Message Passing Interface on IBM SP2 to highlight the speedup that can be achieved over the serial implementation of GP. We also show how data parallelism can be used to further speed up fitness evaluation and hence the execution of the GP paradigm for multicategory pat tern classification. We conclude that GP can be applied to n-category pattern classification and its potential lies in its simplicity and scope for parallel implementation. The GP classifier developed in this thesis can be looked upon as an addition to the earlier statistical, neural and fuzzy approaches to multicategory pattern classification.

Nature has created complex biological structures that exhibit intelligent behaviour through an evolutionary process. Thus, intelligence and evolution are intimately connected. This has inspired evolutionary computation (EC) that simulates the evolutionary process to develop powerful techniques such as genetic algorithms (GAs), genetic programming (GP), evolutionary strategies (ES) and evolutionary programming (EP) to solve real-world problems in learning, control, optimization and classification. GP discovers the relationship among data and expresses it as a LISP-S expression i.e., a computer program. Thus the goal of program discovery as a solution for a problem is addressed by GP in the framework of evolutionary computation. In this thesis, we address for the first time the problem of applying GP to mu1ticategory pattern classification. In supervised pattern classification, an input vector of m dimensions is mapped onto one of the n classes. It has a number of application areas such as remote sensing, medical diagnosis etc., A supervised classifier is developed by using a training set that contains representative samples of various classes present in the application. Supervised classification has been done earlier with maximum likelihood classifier: neural networks and fuzzy logic. The major considerations in applying GP to pattern classification are listed below: (i) GP-based techniques are data distribution-free i.e., no a priori knowledge is needed abut the statistical distribution of the data or no assumption such as normal distribution for data needs to be made as in MLC. (ii) GP can directly operate on the data in its original form. (iii) GP can detect the underlying but unknown relationship that mists among data and express it as a mathematical LISP S-expression. The generated LISP S-expressions can be directly used in the application environment. (iv) GP can either discover the most important discriminating features of a class during evolution or it requires minor post-processing of the LISP-S expression to discover the discriminant features. In a neural network, the knowledge learned by the neural network about the data distributions is embedded in the interconnection weights and it requires considerable amount of post-processing of the weights to understand the decision of the neural network. In 2-category pattern classification, a single GP expression is evolved as a discriminant function. The output of the GP expression can be +l for samples of one class and -1 for samples of the other class. When the GP paradigm is applied to an n-class problem, the following questions arise: Ql. As a typical GP expression returns a value (+l or -1) for a 2-class problem, how does one apply GP for the n-class pattern classification problem? Q2. What should be the fitness function during evolution of the GP expressions? Q3. How does the choice of a function set affect the performance of GP-based classification? Q4. How should training sets be created for evaluating fitness during the evolution of GP classifier expressions? Q5. How does one improve learning of the underlying data distributions in a GP framework? Q6. How should conflict resolution be handled before assigning a class to the input feature vector? Q7. How does GP compare with other classifiers for an n-class pattern classification problem? The research described here seeks to answer these questions. We show that GP can be applied to an n-category pattern classification problem by considering it as n 2-class problems. The suitability of this approach is demonstrated by considering a real-world problem based on remotely sensed satellite images and Fisher's Iris data set. In a 2-class problem, simple thresholding is sufficient for a discriminant function to divide the feature space into two regions. This means that one genetic programming classifier expression (GPCE) is sufficient to say whether or not the given input feature vector belongs to that class; i.e., the GP expression returns a value (+1 or -1). As the n-class problem is formulated as n 2-class problems, n GPCEs are evolved. Hence, n GPCE specific training sets are needed to evolve these n GPCEs. For the sake of illustration, consider a 5-class pat tern classification problem. Let n, be the number of samples that belong to class j, and N, be the number of samples that do not belong to class j, (j = 1,..., 5). Thus, N1=n2+n3+n4+n5 N2=n1+n3+n4+n5 N3=n1+n2+n4+n5 N4=n1+n2+n3+n5 N5=n1+n2+n3+n4 Thus, When the five class problem is formulated as five 2-class problems. we need five GPCEs as discriminant functions to resolve between n1 and N1, n2 and N2, n3 and N3, n4 and N4 and lastly n5 and N5. Each of these five 2-class problems is handled as a separate 2-class problem with simple thresholding. Thus, GPCE# l resolves between samples of class# l and the remaining n - 1 classes. A training set is needed to evaluate the fitness of GPCE during its evolution. If we directly create the training set, it leads to skewness (as n1 < N1). To overcome the skewness, an interleaved data format is proposed for the training set of a GPCE. For example, in the training set of GPCE# l, samples of class# l are placed alternately between samples of the remaining n - 1 classes. Thus, the interleaved data format is an artifact to create a balanced training set. Conventionally, all the samples of a training set are fed to evaluate the fitness of every member of the population in each generation. We call this "global" learning 3s GP tries to learn the entire training set at every stage of the evolution. We have introduced incremental learning to simplify the task of learning for the GP paradigm. A subset of the training set is fed and the size of the subset is gradually increased over time to cover the entire training data. The basic motivation for incremental learning is to improve learning during evolution as it is easier to learn a smaller task and then to progress from a smaller task to a bigger task. Experimental results are presented to show that the interleaved data format and incremental learning improve the performance of the GP classifier. We also show that the GPCEs evolved with an arithmetic function set are able to track variation in the input better than GPCEs evolved with function sets containing logical and nonlinear elements. Hence, we have used arithmetic function set, incremental learning, and interleaved data format to evolve GPCEs in our simulations. AS each GPCE is trained to recognize samples belonging to its own class and reject samples belonging to other classes a strength of association measure is associated with each GPCE to indicate the degree to which it can recognize samples belonging to its own class. The strength of association measures are used for assigning a class to an input feature vector. To reduce misclassification of samples, we also show how heuristic rules can be generated in the GP framework unlike in either MLC or the neural network classifier. We have also studied the scalability and generalizing ability of the GP classifier by varying the number of classes. We also analyse the performance of the GP classifier by considering the well-known Iris data set. We compare the performance of classification rules generated from the GP classifier with those generated from neural network classifier, (24.5 method and fuzzy classifier for the Iris data set. We show that the performance of GP is comparable to other classifiers for the Iris data set. We notice that the classification rules can be generated with very little post-processing and they are very similar to the rules generated from the neural network and C4.5 for the Iris data set. Incremental learning influences the number of generations available for GP to learn the data distribution of classes whose d is -1 in the interleaved data format. This is because the samples belonging to the true class (desired output d is +1) are alternately placed between samples belonging to other classes i.e., they are repeated to balance the training set in the interleaved data format. For example, in the evolution of GPCE for class# l, the fitness function can be fed initially with samples of class#:! and subsequently with the samples of class#3, class#4 and class#. So in the evaluation of the fitness function, the samples of class#kt5 will not be present when the samples of class#2 are present in the initial stages. However, in the later stages of evolution, when samples of class#5 are fed, the fitness function will utilize the samples of both class#2 and class#5. As learning in evolutionary computation is guided by the evaluation of the fitness function, GPCE# l gets lesser number of generations to learn how to reject data of class#5 as compared to the data of class#2. This is because the termination criterion (i.e., the maximum number of generations) is defined a priori. It is clear that there are (n-l)! Ways of ordering the samples of classes whose d is -1 in the interleaved data format. Hence a heuristic is presented to determine a possible order to feed data of different classes for the GPCEs evolved with incremental learning and interleaved data format. The heuristic computes an overlap index for each class based on its spatial spread and distribution of data in the region of overlap with respect to other classes in each feature. The heuristic determines the order in which classes whose desired output d is –1 should be placed in each GPCE-specific training set for the interleaved data format. This ensures that GP gets more number of generations to learn about the data distribution of a class with higher overlap index than a class with lower overlap index. The ability of the GP classifier to learn the data distributions depends upon the number of classes and the spatial spread of data. As the number of classes increases, the GP classifier finds it difficult to resolve between classes. So there is a need to partition the feature space and identify subspaces with reduced number of classes. The basic objective is to divide the feature space into subspaces and hence the data set that contains representative samples of n classes into subdata sets corresponding to the subspaces of the feature space, so that some of the subdata sets/spaces can have data belonging to only p classes (p < n). The GP classifier is then evolved independently for the subdata sets/spaces of the feature space. This results in localized learning as the GP classifier has to learn the data distribution in only a subspace of the feature space rather than in the entire feature space. By integrating the GP classifier with feature space partitioning (FSP), we improve classification accuracy due to localized learning. Although serial computers have increased steadily in their performance, the quest for parallel implementation of a given task has continued to be of interest in any computationally intensive task since parallel implementation leads to a faster execution than a serial implementation As fitness evaluation, selection strategy and population structures are used to evolve a solution in GP, there is scope for a parallel implementation of GP classifier. We have studied distributed GP and massively parallel GP for our approach to GP-based multicategory pattern classification. We present experimental results for distributed GP with Message Passing Interface on IBM SP2 to highlight the speedup that can be achieved over the serial implementation of GP. We also show how data parallelism can be used to further speed up fitness evaluation and hence the execution of the GP paradigm for multicategory pat tern classification. We conclude that GP can be applied to n-category pattern classification and its potential lies in its simplicity and scope for parallel implementation. The GP classifier developed in this thesis can be looked upon as an addition to the earlier statistical, neural and fuzzy approaches to multicategory pattern classification.

Parmi les diverses modalités biométriques qui permettent l'identification des personnes, l'iris est considéré comme très fiable, avec un taux d'erreur remarquablement faible. Toutefois, ce niveau élevé de performances est obtenu en contrôlant la qualité des images acquises et en imposant de fortes contraintes à la personne (être statique et à proximité de la caméra). Cependant, dans de nombreuses applications de sécurité comme les contrôles d'accès, ces contraintes ne sont plus adaptées. Les images résultantes souffrent alors de diverses dégradations (manque de résolution, artefacts...) qui affectent négativement les taux de reconnaissance. Pour contourner ce problème, il est possible d’exploiter la redondance de l’information découlant de la disponibilité de plusieurs images du même œil dans la séquence enregistrée. Cette thèse se concentre sur la façon de fusionner ces informations, afin d'améliorer les performances. Dans la littérature, diverses méthodes de fusion ont été proposées. Cependant, elles s’accordent sur le fait que la qualité des images utilisées dans la fusion est un facteur crucial pour sa réussite. Plusieurs facteurs de qualité doivent être pris en considération et différentes méthodes ont été proposées pour les quantifier. Ces mesures de qualité sont généralement combinées pour obtenir une valeur unique et globale. Cependant, il n'existe pas de méthode de combinaison universelle et des connaissances a priori doivent être utilisées, ce qui rend le problème non trivial. Pour faire face à ces limites, nous proposons une nouvelle manière de mesurer et d'intégrer des mesures de qualité dans un schéma de fusion d'images, basé sur une approche de super-résolution. Cette stratégie permet de remédier à deux problèmes courants en reconnaissance par l'iris: le manque de résolution et la présence d’artefacts dans les images d'iris. La première partie de la thèse consiste en l’élaboration d’une mesure de qualité pertinente pour quantifier la qualité d’image d’iris. Elle repose sur une mesure statistique locale de la texture de l’iris grâce à un modèle de mélange de Gaussienne. L'intérêt de notre mesure est 1) sa simplicité, 2) son calcul ne nécessite pas d'identifier a priori les types de dégradations, 3) son unicité, évitant ainsi l’estimation de plusieurs facteurs de qualité et un schéma de combinaison associé et 4) sa capacité à prendre en compte la qualité intrinsèque des images mais aussi, et surtout, les défauts liés à une mauvaise segmentation de la zone d’iris. Dans la deuxième partie de la thèse, nous proposons de nouvelles approches de fusion basées sur des mesures de qualité. Tout d’abord, notre métrique est utilisée comme une mesure de qualité globale de deux façons différentes: 1) comme outil de sélection pour détecter les meilleures images de la séquence et 2) comme facteur de pondération au niveau pixel dans le schéma de super-résolution pour donner plus d'importance aux images de bonnes qualités. Puis, profitant du caractère local de notre mesure de qualité, nous proposons un schéma de fusion original basé sur une pondération locale au niveau pixel, permettant ainsi de prendre en compte le fait que les dégradations peuvent varier d’une sous partie à une autre. Ainsi, les zones de bonne qualité contribueront davantage à la reconstruction de l'image fusionnée que les zones présentant des artéfacts. Par conséquent, l'image résultante sera de meilleure qualité et pourra donc permettre d'assurer de meilleures performances en reconnaissance. L'efficacité des approches proposées est démontrée sur plusieurs bases de données couramment utilisées: MBGC, Casia-Iris-Thousand et QFIRE à trois distances différentes. Nous étudions séparément l'amélioration apportée par la super-résolution, la qualité globale, puis locale dans le processus de fusion. Les résultats montrent une amélioration importante apportée par l'utilisation de la qualité globale, amélioration qui est encore augmentée en utilisant la qualité locale<br>Among the large number of biometric modalities, iris is considered as a very reliable biometrics with a remarkably low error rate. The excellent performance of iris recognition systems are obtained by controlling the quality of the captured images and by imposing certain constraints on users, such as standing at a close fixed distance from the camera. However, in many real-world applications such as control access and airport boarding these constraints are no longer suitable. In such non ideal conditions, the resulting iris images suffer from diverse degradations which have a negative impact on the recognition rate. One way to try to circumvent this bad situation is to use some redundancy arising from the availability of several images of the same eye in the recorded sequence. Therefore, this thesis focuses on how to fuse the information available in the sequence in order to improve the performance. In the literature, diverse schemes of fusion have been proposed. However, they agree on the fact that the quality of the used images in the fusion process is an important factor for its success in increasing the recognition rate. Therefore, researchers concentrated their efforts in the estimation of image quality to weight each image in the fusion process according to its quality. There are various iris quality factors to be considered and diverse methods have been proposed for quantifying these criteria. These quality measures are generally combined to one unique value: a global quality. However, there is no universal combination scheme to do so and some a priori knowledge has to be inserted, which is not a trivial task. To deal with these drawbacks, in this thesis we propose of a novel way of measuring and integrating quality measures in a super-resolution approach, aiming at improving the performance. This strategy can handle two types of issues for iris recognition: the lack of resolution and the presence of various artifacts in the captured iris images. The first part of the doctoral work consists in elaborating a relevant quality metric able to quantify locally the quality of the iris images. Our measure relies on a Gaussian Mixture Model estimation of clean iris texture distribution. The interest of our quality measure is 1) its simplicity, 2) its computation does not require identifying in advance the type of degradations that can occur in the iris image, 3) its uniqueness, avoiding thus the computation of several quality metrics and associated combination rule and 4) its ability to measure the intrinsic quality and to specially detect segmentation errors. In the second part of the thesis, we propose two novel quality-based fusion schemes. Firstly, we suggest using our quality metric as a global measure in the fusion process in two ways: as a selection tool for detecting the best images and as a weighting factor at the pixel-level in the super-resolution scheme. In the last case, the contribution of each image of the sequence in final fused image will only depend on its overall quality. Secondly, taking advantage of the localness of our quality measure, we propose an original fusion scheme based on a local weighting at the pixel-level, allowing us to take into account the fact that degradations can be different in diverse parts of the iris image. This means that regions free from occlusions will contribute more in the image reconstruction than regions with artefacts. Thus, the quality of the fused image will be optimized in order to improve the performance. The effectiveness of the proposed approaches is shown on several databases commonly used: MBGC, Casia-Iris-Thousand and QFIRE at three different distances: 5, 7 and 11 feet. We separately investigate the improvement brought by the super-resolution, the global quality and the local quality in the fusion process. In particular, the results show the important improvement brought by the use of the global quality, improvement that is even increased using the local quality

Técnicas biométricas procuraram identificar usuários pela textura da íris, impressão digital, traços faciais, entre outros. A íris humana apresenta características de textura que a classificam como uma peculiaridade biométrica de grande poder de discriminação no reconhecimento de pessoas. O objetivo deste trabalho é avaliar a eficiência de uma nova metodologia de análise de texturas em desenvolvimento no LAVI (Laboratório de Visão Computacional da EESC-USP) na identificação de indivíduos por meio da textura de sua íris. A metodologia denomina-se Local Fuzzy Pattern e tem sido utilizada com excelente desempenho com texturas gerais, naturais e artificiais. Este documento detalha as técnicas utilizadas para extração e normalização da textura da íris, a utilização e os resultados obtidos com o método Local Fuzzy Pattern aplicado à classificação biométrica da íris humana. Os resultados obtidos apresentam sensibilidade de até 99,7516% com a aplicação da metodologia proposta em bancos de imagens de íris humana disponíveis na internet demonstram a viabilidade da técnica proposta.<br>Biometric techniques sought to identify users by the texture of the iris, fingerprint, facial features, among others. The human iris have texture characteristics that rank it as a powerful biometric peculiarity on human recognition. The objective of this masters proposal is to investigate the efficiency of a new methodology of iris texture analysis currently in development in LAVI (Laboratório de Visão Computacional da EESC-USP). The methodology is called LFP (Local Fuzzy Pattern) and has been used with excellent overall performance on artificial and natural textures. This document details the techniques used for the extraction and normalization of the iris texture, the use and results of the local fuzzy pattern method applied to biometric classification of the human eye. The results show a sensibility of value up to 99.7516%, obtained by applying the proposed methodology on human iris photos from image database available on the internet does showing the viability of the technique.

Made available in DSpace on 2016-06-02T19:05:57Z (GMT). No. of bitstreams: 1 4427.pdf: 8518956 bytes, checksum: 0179ef9750c36082852192a44b3e6834 (MD5) Previous issue date: 2012-06-14<br>Financiadora de Estudos e Projetos<br>The identification of humans by their iris structure has been explored since 1993, when the first algorithm was made available by John Daugman. Since then, iris recognition systems are widely used for access control of several kinds of environments. Such systems typically requires the user´s cooperation, appropriate lighting conditions, and images obtained in the infra-red band. Dynamic methods for biometric identification has been the subject of studies in the past few years, including iris recognition in non-cooperative environments. This paper proposes a pre-processing methodology to enable iris images classification taken in a noncooperative setting, from users at a certain distance, or while moving. The methodology aims to select images from the visible band containing an acceptable level of noise, and as such being suitable to apply the classification algorithms. Experimental results have shown that images with up to 40% of noise can still be used, suggesting the methodology may be useful as an aid to implement iris recognition systems at distance or in motion.<br>A identificação de seres humanos pela estrutura da íris vem sendo explorada desde 1993, quando foi disponibilizado o primeiro algoritmo por John Daugman. Desde então, os sistemas de reconhecimento de íris são amplamente utilizados para o controle de acesso de diversas aplicações. Tais sistemas normalmente, requerem a cooperação do usuário, condições de iluminações adequadas, e imagens obtidas na banda infravermelha. Métodos dinâmicos para identificação biométrica tem sido objeto de estudo nos últimos anos, incluindo o reconhecimento de íris em ambientes não cooperativos. Este trabalho propõe uma metodologia de pré-processamento da imagens da íris para classificação de amostras capturadas de forma não cooperativa, a uma certa distância, ou em movimento pelo usuário. A metodologia visa selecionar imagens a partir da banda visível contendo um nível de ruído aceitável, de forma que possa ser eficaz na aplicação dos algoritmos de classificação. Resultados experimentais demostraram que imagens com até 40% de ruído podem ainda ser utilizadas, sugerindo a utilização da metodologia como um auxílio para implementação de sistemas de reconhecimento de íris à distância ou em movimento.

The Irish clachan is a settlement form, a grouping of houses and related out-buildings, that has been associated with the 'Rundale' agricultural system. Study of it has proceeded largely within the discipline of Geography, and has been coordinated primarily by questions such as those of the antiquity and distribution of the type. The morphology of the built form has been commented on, but in a limited and highly circumscribed way; to date it has received little sustained attention. This is the focus of the present study - the spatial formation of the clusters, their 'architectural' quality, and specifically how that was and is interpreted and understood. Following an introductory critical review of recent academic treatment of the subject, a brief history is sketched of ways in which the clusters have been described. Close attention is paid to the categories which relate to their spatial qualities, categories which, it is recognized, passed over into the documents of the human sciences and which problematize the latters' project of understanding. It is thus argued, with reference to Heidegger's thinking, that scholarship on the clusters has conceptually mis-sited them. The hermeneutic position of the investigator has not been thematized. By situating the commentaries on native settlement within the development of the expansive, and far from uncommitted, discourse on Ireland the operation of the notion of disorder is examined. Here it is seen to be structurally related to a series of privileged and abject categories strategically organized around a master duality of civility and savagery. The suggestion that geometrically structured space, as the space of authority, was importantly linked with the privileged categories is investigated and the relationship between the clusters and the categories of the 'monstrous' and the 'grotesque' considered.

During the twelfth and thirteenth-centuries, Ireland experienced a large-scale urbanization movement, initiated as part of the Anglo-Norman conquest and colonization of the island. As part of this process, old settlements were re-modelled and promoted; and an extensive network of new towns were planted across the Irish medieval landscape. This dissertation examines the development of this colonial urbanization movement with particular reference to the urban planning aspects of the process. Volume I, considers the origins, influences, and ideals of the Anglo-Norman town builders, as well as the morphogenetic, spatial and distributive characteristics of their endeavors. In addition, the current level of scholarship on the subject is highlighted and discussed. Volume II, focuses on a series of typical Anglo-Norman town foundations, where the origins, plan and burgage development patterns are explored in depth, through the techniques of plan and metrological analysis.

This thesis examines how a new media technology becomes entrenched into the fabric of society – in particular how the mobile phone was incorporated into the existing communications landscape of a local voluntary community group: an Irish sports club. In the past, face-to-face interaction formed the basis of all social relations and strong local collectives were seen to provide a positive and supportive social environment, generating strong social capital. Today’s mediated communication enables the ‘networked individual’ who can choose when, where and with whom they share their lives. This has implications for the persistence and strength of local associations. Writers like Putnam (2000) have expressed concerns about the attenuation of local communities. However, others propose that communications technologies can provide new additional ways for individuals to link with each other in a ‘glocalised’ society (Hampton and Wellman, 2003) and this has the potential to overcome some of the limitations of the communicators not being in the same physical space. This thesis critically examines the applicability of these partly competing theses in the period in which mobile telephony became widely embedded in Irish society. Through 21 detailed interviews and a survey of 57 players, administrators and supporters, the study examines the choices made by club members in adopting and using the mobile phone. It further explores the changes they have made in their communication patterns and considers the implications of these for the cohesion and persistence of the community group as an entity and also the social capital it engenders. The study draws upon existing theories of human-technology interaction, in particular the Social Shaping of Technology perspective (Williams and Edge, 1996), to examine how club members weave their phone use into their everyday practices. Silverstone and Haddon’s (1996) Domestication approach, with its steps of appropriation, objectification and incorporation provide the detailed framework in mapping out this process. The study findings reveal that community members have all adopted the mobile phone and are heavy users of both text and voice calls. Membership of the sports club has eased the adoption process by providing examples of the artefact in use and a supporting environment when problems arise. Although use is now universal and intensive, there was a differential appropriation of the phone, with male club members being the first purchasers and females often being brought into the circle of users through a gifted or handed-down model. Users have devised strategies to manage their multiple overlapping sets of social relationship. They report that their use of technology has enabled a widening of their social circle while also bringing it closer, literally at the touch of a button. Contrary to the expectations of those analysts and policymakers who have foreseen technology causing local engagement to diminish, the clubs in my study have endured and are thriving; the social capital of their members is still strong and growing. This informs a critical reappraisal of such theories of community attenuation and the policies they have engendered.

The Irish Ice Sheet has been an active area of research since the mid 19th century. Traditional approaches to reconstructing the Irish Ice Sheet have used the landform and sediment record to define its extent and formulate models of its geometry through the Pleistocene. Subglacial bedforms dominate the landform record and till is the dominant sediment. The bedforms demonstrate a complex ice flow history, however recent research has indicated that ice flow directions interpreted from subglacial bedform geomorphology could be misleading. It is therefore appropriate to consider other methods of investigating ice flow patterns that are independent of geomorphology. As the bedforms are primarily composed of till, an alternative approach to determining flow directions is to establish the provenance of the till using geochemistry. High density regional soil geochemical data sets comprising> 1 0,000 sample points are available for the north of Ireland. This thesis applies a multi variate statistical approach to the soil geochemistry to establish till provenance in the northern sector of the Irish Ice Sheet. Till provenance is determined for the entire region and also investigated in a range of palaeoglaciological contexts. Results are compared to recent models of flow history for the Irish Ice Sheet. Till provenance is also applied to determine the degree of glacial dispersal of gold mineralisation from known host regions in the north of Ireland. The thesis demonstrates that till in this sector of the Irish Ice Sheet shows a very strong geochemical link to local bedrock composition, with sharp geochemical transitions observed at lithological boundaries. However, subglacial transport and clear dispersal patterns are detected particularly at flow set scale in a range of palaeoglaciological contexts. Geochemical evidence is found for flow directions for each of the major phases of the history of the last Irish Ice Sheet and this evidence compliments existing models. No unequivocal glacial dispersal patterns are detected for gold, though the prospectivity of currently unexplored areas is highlighted. The research demonstrates that soil geochemistry makes a significant contribution to Quaternary research in Ireland which has not been previously recognised.

Karabaglar is a rural area where agricultural community lives and which locates in the vicinity of Mugla town. It was registered as third grade natural site that must be preserved according to its rural character, natural and cultural assets and rural life. Property pattern consisting of private ownerships (yurts), specific road network that emerges from irims and kesiks, traditional houses, variety of vegetation, abundant water, self-sufficient agricultural production, and traditional life style are the main features that create and shape Karabaglar. Kesiks, irims, kabaliks, and yurts are the major man-made components, which are unique to Karabaglar, and these characteristics conform to the natural landscape structure. Urban growth, interventions due to misuse of lands, new housing demands of urban residents in Karabaglar resulted in urban pressure on the area. This situation puts forth the conservation necessity of Karabaglar. This research analyzes the speculative housing development in Karabaglar, identifies the type of interventions and their physical, social, economic and environmental effects on Karabaglar<br>in addition, develops proposals for sustainability of the traditional pattern of Karabaglar.

This thesis analyses how an increase in bullying awareness can be produced in children aged 9-12 years old playing an iPad serious game called Attitude Game. The project is conducted in collaboration with the company IUS Innovation, where two prototypes were developed by the author of the study, one with mechanics empowering not to bully and the other with mirror mechanics empowering to bully other children. The experiment was conducted with 29 children from a sports club in Göteborg.  A pre-measurement of empathy and a pre-test-post-test measurement of attitude towards bullied children were used to gather data, together with a post-interview during the experiment. Students were assigned to 3 groups, based on a random selection and previous knowledge about the game. The result of the experiment displayed that there was no statistical difference between the change in attitude towards bullied children regarding the empathy level and the prototype tested. The post-interviews displayed that all the participants understood the serious purpose of the game<br><p>In case of further contact, you can also write to the private email of the author: carlosglesdiaz@gmail.com</p>

碩士<br>國立中正大學<br>電機工程研究所<br>103<br>In today's advanced information age, security plays a very important position. The popular PIN (Personal Identification Number) Identification is usually lost because of negligence, so " biometric identification system " which has higher security will be used. The people does not need to memory a long list of password. Iris recognition is one of the most accurate biometric identification methods. Iris of each person is different, and its feature is not easily changed and has high accuracy. In the thesis, firstly we capture the iris image, and then using pre-processing algorithm to remove noise and complete the location to be the iris features ROI. Then it lets the features turn into iris code. In order to verify the correctness of the system, two databases are used. The first one is CASIA iris image database which is provided by the Chinese Academy of Sciences Institute of Automation, the other is CCU-database iris image database which is provided by Taiwan National Chung Cheng University. The matching rates of our sample are respectively 97.36% and 92.40%.

The development of accurate and reliable security systems is a matter of wide interest, and in this context biometrics is seen as a highly effective automatic mechanism for personal identification. Among biometric technologies, IrisCode developed by Daugman in 1993 is regarded as a highly accurate approach, being able to support real-time personal identification of large databases. Since 1993, on the top of IrisCode, different coding methods have been proposed for iris and fingerprint identification. In this research, I extend and generalize IrisCode for real-time secure palmprint identification. PalmCode, the first coding method for palmprint identification developed by me in 2002, directly applied IrisCode to extract phase information of palmprints as features. However, I observe that the PalmCodes from the different palms are similar, having many 45o streaks. Such structural similarities in the PalmCodes of different palms would reduce the individuality of PalmCodes and the performance of palmprint identification systems. To reduce the correlation between PalmCodes, in this thesis, I employ multiple elliptical Gabor filters with different orientations to compute different PalmCodes and merge them to produce a single feature, called Fusion Code. Experimental results demonstrate that Fusion Code performs better than PalmCode. Based on the results of Fusion Code, I further identify that the orientation fields of palmprints are powerful features. Consequently, Competitive Code, which uses real parts of six Gabor filters to estimate the orientation fields, is developed. To embed the properties of IrisCode, such as high speed matching, in Competitive Code, a novel coding scheme and a bitwise angular distance are proposed. Experimental results demonstrate that Competitive Code is much more effective than other palmprint algorithms. Although many coding methods have been developed based on IrisCode for iris and palmprint identification, we lack a detailed analysis of IrisCode. One of the aims of this research is to provide such analysis as a way of better understanding IrisCode, extending the coarse phase representation to a precise phase representation, and uncovering the relationship between IrisCode and other coding methods. This analysis demonstrates that IrisCode is a clustering process with four prototypes; the locus of a Gabor function is a two-dimensional ellipse with respect to a phase parameter and the bitwise hamming distance can be regarded as a bitwise angular distance. In this analysis, I also point out that the theoretical evidence of the imposter binomial distribution of IrisCode is incomplete. I use this analysis to develop a precise phase representation which can enhance iris recognition accuracy and to relate IrisCode and other coding methods. By making use of this analysis, principal component analysis and simulated annealing, near optimal filters for palmprint identification are sought. The near optimal filters perform better than Competitive Code in term of d’ index. Identical twins having the closest genetics-based relationship are expected to have maximum similarity in their biometrics. Classifying identical twins is a challenging problem for some automatic biometric systems. Palmprint has been studied for personal identification for many years. However, genetically identical palmprints have not been studied. I systemically examine Competitive Code on genetically identical palmprints for automatic personal identification and to uncover the genetically related palmprint features. The experimental results show that the three principal lines and some portions of weak lines are genetically related features but our palms still contain rich genetically unrelated features for classifying identical twins. As biometric systems are vulnerable to replay, database and brute-force attacks, such potential attacks must be analyzed before they are massively deployed in security systems. I propose projected multinomial distribution for studying the probability of successfully using brute-force attacks to break into a palmprint system based on Competitive Code. The proposed model indicates that it is computationally infeasible to break into the palmprint system using brute-force attacks. In addition to brute-force attacks, I address the other three security issues: template re-issuances, also called cancellable biometrics, replay attacks, and database attacks. A random orientation filter bank (ROFB) is used to generate cancellable Competitive Codes for templates re-issuances. Secret messages are hidden in templates to prevent replay and database attacks. This technique can be regarded as template watermarking. A series of analyses is provided to evaluate the security levels of the measures.

A biometric system of identification and authentication provides automatic recognition of an individual based on certain unique features or characteristics possessed by that individual. Iris recognition is a biometric identification method that uses pattern recognition on the images of the iris of an individual. Iris recognition is considered as one of the most accurate biometric methods available owing to the unique epigenetic patterns of the iris. In this project, we have developed a system that can recognize human iris patterns and an analysis of the results is done. A hybrid mechanism has been used for implementation of the system. Iris localization is done by amalgamating the Canny Edge Detection scheme and Circular Hough Transform. The iris images are then normalized so as to transform the iris region to have fixed dimensions in order to allow comparisons. Feature encoding has been used to extract the most discriminating features of the iris and is done using a modification of Gabor wavelets. And finally the biometric templates are compared using Hamming Distance which tells us whether the two iris images are same or not.

博士<br>國立暨南國際大學<br>資訊工程學系<br>98<br>In this dissertation, we study the influences of both the non-orthogonal imaging condition and the nonlinear deformation of iris patterns on the accuracy of iris recognition, and we also propose effective methods to deal with these two topics. First, we propose a non-orthogonal view iris recognition system which comprises a new iris imaging module, an iris segmentation module, an iris feature extraction module and a classification module. A dual-CCD camera is developed to capture four-spectral (red, green, blue and near-infrared) iris images which contain useful information for simplifying the iris segmentation task. An intelligent RANSAC iris segmentation method is proposed to robustly detect iris boundaries in a four-spectral iris image. In order to match iris images acquired at different off-axis angles, we propose a circle rectification method to reduce the off-axis iris distortion. The rectification parameters are estimated using the detected elliptical pupillary boundary. Furthermore, we propose a novel iris descriptor which characterizes an iris pattern with multi-scale step/ridge edge-type maps. The edge-type maps are extracted with the derivative of Gaussian and the Laplacian of Gaussian filters. The iris pattern classification is accomplished by edge-type matching which can be understood intuitively with the concept of classifier ensembles. Experimental results show that the equal error rate of our approach is only 0.04% when recognizing iris images acquired at different off-axis angles within ±30◦. Additionally, a nonlinear iris normalization method is proposed. This method can handle iris deformation due to myosis/mydriasis. In order to prove the feasibility of our method, another iris imaging system is constructed. This system includes a computer controllable current source for driving a blue LED array, which is used to capture iris deformation images at different light intensity levels. Experimental result shows that our proposed method outperforms the traditional linear normalization method. The equal error rates of our and the traditional linear normalization method are 0.95% and 2.76%, respectively.

碩士<br>國立暨南國際大學<br>資訊工程學系<br>99<br>Nowadays, automatic iris recognition systems can achieve excellent performance when the pupil sizes of both the enrolled and the test iris images are approximately the same. How- ever, whenthepupilsizechangesdramaticallyduetoeitherilluminationvariationoremotion stimuli, the equal error rate (EER) will increase because of strong nonlinear iris deformation. In this thesis, we focus on the normalization method for compensating the nonlinear iris de- formation effect. Two nonlinear normalization approaches, i.e., the information equalization (IE) method and the dynamic programming (DP) method, are proposed in this study. We also implement Daugman’s rubber sheet model and Yuan’s nonlinear normalization model to be compared with the two proposed methods. Furthermore, a special iris image database called IrisDeform database is constructed with 228 classes of iris patterns from 114 subjects. Each class in the IrisDeform database contains 160 iris patterns acquired at two different sessions with 16 different illumination intensity levels and five iris images for each intensity level. The IrisDeform database is useful for researchers working on the nonlinear normalization problem of iris patterns. Two experiments are conducted to test the nonlinear normalization algorithms. In the first experiment, we enroll the iris images by turns and use the remaining images to method and Yuan’s nonlinear normalization method are 3.21% and 3.51%, respec- tively, whereas the EERs of our IE and DP methods are 4.25% and 2.26%, respectively. The second experiment aims to test the performance of different normalization methods at the extreme cases. The enrolled and the test iris images are acquired at the darkest and the brightest illumination intensity, respectively. The experimental results show that the EERs of both the methods of Daugman and Yuan degraded to 7.21% and 8.05%, respectively, due to strong iris deformation. Conversely, the EER of the DP method only increases to 6.93%. The two experiments show that the proposed DP method is more tolerable to large nonlinear deformations of iris patterns.

Today Andrew Marvell's poetry is thought to offer a window onto mid-seventeenth-century English literature and culture, yet scholars find the poet's richly allusive early works puzzling: we often do not know what prompted these compositions, or how to interpret them. Marvell probably wrote much of his early verse in 1651-1652 while working as a tutor at the Fairfax family's Yorkshire estate, Nun Appleton. Four approaches to Marvell's major early work, the estate poem Upon Appleton House, help to clarify the poet's methodology, the Yorkshire cultural and landscape milieus of his 1651-1652 poems, the prominent family for which he worked, and the pedagogic content of the poem itself. In the first approach, textual analysis and pattern-tracing reveal that Marvell developed Upon Appleton House from short poetic studies in Latin and English, and reveal too some ways in which Marvell represented his employer, Thomas Fairfax; his student, Mary Fairfax; and himself, as tutor-poet persona. Next, research on central Yorkshire's historical geography and lore and especially on Fairfax family lands helps explicate Upon Appleton House and shows that Marvell himself was a researcher and close observer of the outdoors. Third, information about the career and retirement of Thomas Fairfax--who in 1650 was nominally Interregnum England's highest-ranking leader--partially demystifies both Fairfax's retirement motives and Marvell's poem. A final approach analyzes Upon Appleton House as a poem for the instruction of thirteen-year-old Mary Fairfax. Marvell apparently drew on ideas from advice-to-a-prince poems, education manuals, puritan theology, and other sources to prepare Mary Fairfax for her future roles as Protestant heiress, dynastic perpetuator, and "natural ruler." Moreover, Marvell lyrically transformed the lands she would inherit into a medium for learning. Each approach to Upon Appleton House includes attention to literary and visual arts' traditions and to Marvell's evolution as a poet. Together, the four approaches go far toward explaining Marvell's 1651-1652 compositional chronology and self-presentation, his descriptions of nature and Yorkshire landscapes, his praise and instruction of Fairfax family members, and his evocations of post-civil-war England.