Dissertationen zum Thema „Re-identification and image enhancement“

Background: Video surveillance is a growing area where it can help with deterring crime, support investigation or to help gather statistics. These are just some areas where video surveillance can aid society. However, there is an improvement that could increase the efficiency of video surveillance by introducing tracking. More specifically, tracking between cameras in a network. Automating this process could reduce the need for humans to monitor and review since the tracking can track and inform the relevant people on its own. This has a wide array of usability areas, such as forensic investigation, crime alerting, or tracking down people who have disappeared. Objectives: What we want to investigate is the common setup of real-time multi-target multi-camera tracking (MTMCT) systems. Next up, we want to investigate how the components in an MTMCT system affect each other and the complete system. Lastly, we want to see how image enhancement can affect the MTMCT. Methods: To achieve our objectives, we have conducted a systematic literature review to gather information. Using the information, we implemented an MTMCT system where we evaluated the components to see how they interact in the complete system. Lastly, we implemented two image enhancement techniques to see how they affect the MTMCT. Results: As we have discovered, most often, MTMCT is constructed using a detection for discovering object, tracking to keep track of the objects in a single camera and a re-identification method to ensure that objects across cameras have the same ID. The different components have quite a considerable effect on each other where they can sabotage and improve each other. An example could be that the quality of the bounding boxes affect the data which re-identification can extract. We discovered that the image enhancement we used did not introduce any significant improvement. Conclusions: The most common structure for MTMCT are detection, tracking and re-identification. From our finding, we can see that all the component affect each other, but re-identification is the one that is mostly affected by the other components and the image enhancement. The two tested image enhancement techniques could not introduce enough improvement, but other image enhancement could be used to make the MTMCT perform better. The MTMCT system we constructed did not manage to reach real-time.

Resolution is the most important criterion for the clarity of details on an image. Therefore, high resolution images are required in numerous areas. However, obtaining high resolution images has an evident technological cost and the value of these costs change with the quality of used optical systems. Image processing methods are used to obtain high resolution images with low costs. This kind of image improvement is named as superresolution image reconstruction. This thesis focuses on two main titles, one of which is the identification methods of blur parameters, one of the degradation operators, and the stochastic SR image reconstruction methods. The performances of different stochastic SR image reconstruction methods and blur identification methods are shown and compared. Then the identified blur parameters are used in superresolution algorithms and the results are shown.

Nowadays biometric identification is used in a variety of applications-administration, business and even home. Although there are a lot of biometric identifiers, fingerprints are the most widely spread due to their acceptance from the people and the cheap price of the hardware equipment. Fingerprint recognition is a complex image recognition problem and includes algorithms and procedures for image enhancement and binarization, extracting and matching features and sometimes classification. In this work the main approaches in the research area are discussed, demonstrated and tested in a sample application. The demonstration software application is developed by using Verifinger SDK and Microsoft Visual Studio platform. The fingerprint sensor for testing the application is AuthenTec AES2501.

With the increasing demand of intelligent video surveillance systems, person re-identification (re-ID) plays an important role in intelligent video analysis, which aims at matching person images across non-overlapping camera views. It has gained increasing attention in computer vision community. With the advanced deep neural networks, existing methods have achieved promising performance on the widely-used re-ID benchmarks, even outperform the human-level rank-1 matching accuracy. However, most of the research efforts are conducted on the closed-world settings, with large-scale well annotated training data and all the person images are from the same visible modality. As a prerequisite in practical video surveillance application, there is still a large gap between the closed-world research-oriented setting and the practical open-world settings. In this thesis, we try to narrow the gap by studying three important issues in open-world person re-identification, including 1) unsupervised learning with large-scale unlabelled training data; 2) learning robust re-ID model with label corrupted training data and 3) cross-modality visible-thermal person re-identification with multi-modality data. For unsupervised learning with unlabelled training data, we mainly focus on video-based person re-identification, since the video data is usually easily obtained by tracking algorithms and the video sequence provides rich weakly labelled samples by assuming the image frames within the tracked sequence belonging to the same person identity. Following the cross-camera label estimation approach, we formulate the cross-camera label estimation as a one-to-one graph matching problem, and then propose a novel dynamic graph matching framework to estimate cross-camera labels. However, in a practical wild scenario, the unlabelled training data usually cannot satisfy the one-to-one matching constraint, which would result in a large proportion of false positives. To address this issue, we further propose a novel robust anchor embedding method for unsupervised video re-ID. In the proposed method, some anchor sequences are firstly selected to initialize the CNN feature representation. Then a robust anchor embedding method is proposed to measure the relationship between the unlabelled sequences and anchor sequences, which considers both the scalability and efficiency. After that, a top-{dollar}k{dollar} counts label prediction strategy is proposed to predict the labels of unlabelled sequences. With the newly estimated sequences, the CNN representation could be further updated. For robust re-ID model learning with label corrupted training data, we propose a two-stage learning method to handle the label noise. Rather than simply filtering the falsely annotated samples, we propose a joint learning method by simultaneously refining the falsely annotated labels and optimizing the neural networks. To address the limited training samples for each identity, we further propose a novel hard-aware instance re-weighting strategy to fine-tune the learned model, which assigns larger weights to hard samples with correct labels. For cross-modality visible-thermal person re-identification, it addresses an important issue in night-time surveillance applications by matching person images across different modalities. We propose a dual-path network to learn the cross-modality feature representations, which learns the multi-modality sharable feature representations by simultaneously considering the modality discrepancy and commonness. To guide the feature representation learning process, we propose a dual-constrained top-ranking loss, which contains both cross-modality and intra-modality top-ranking constraints to reduce the large cross-modality and intra-modality variations. Besides the open-world person re-identification, we have also studied the unsupervised embedding learning problem for general image classification and retrieval. Motivated by supervised embedding learning, we propose a data augmentation invariant and instance spread-out feature. To learn the feature embedding, we propose a instance feature-based softmax embedding, which optimizes the embedding directly on top of the real-time instance features. It achieves much faster learning speed and better accuracy than existing methods. In short, the major contributions of this thesis are summarized as follows. l A dynamic graph matching framework is proposed to estimate cross-camera labels for unsupervised video-based person re-identification. l A robust anchor embedding method with top-{dollar}k{dollar} counts label prediction is proposed to efficiently estimate the cross-camera labels for unsupervised video-based person re-identification under wild settings. l A two-stage PurifyNet is introduced to handle the label noise problem in person re-identification, which jointly refines the falsely annotated labels and mines hard samples with correct labels. l A dual-constrained top-ranking loss with a dual-path network is proposed for cross-modality visible-thermal person re-identification, which simultaneously addresses the cross-modality and intra-modality variations. l A data augmentation invariant and instance spread-out feature is proposed for unsupervised embedding learning, which directly optimizes the learned embedding on top of real-time instance features with softmax function

Cette thèse concerne la restauration aveugle d’images (formulée comme un problème inverse mal-posé et mal-conditionné), en considérant particulièrement les systèmes SIMO. Dans un premier temps une technique d’identification aveugle de ce système où l’ordre du canal est inconnu (surestimé) est introduite. Nous introduisons d’abord une version simplifiée à coût réduit SCR de la méthode des relations croisées (CR). Ensuite, une version robuste R-SCR basée sur la recherche d’une solution parcimonieuse minimisant la fonction de coût CR est proposée. La restauration d’image est ensuite assurée par une nouvelle approche inspirée des techniques de décodage des signaux 1D et étendue ici aux cas de la restauration d’images en se basant sur une recherche arborescente efficace (algorithme ‘Stack’). Plusieurs améliorations de la méthode ‘Stack’ ont été introduites afin de réduire sa complexité et d’améliorer la qualité de restauration lorsque les images sont fortement bruitées. Ceci en utilisant une technique de régularisation et une approche d’optimisation all-at-once basée sur la descente du gradient qui permet de raffiner l’image estimée et mieux converger vers la solution optimale. Ensuite, les mesures de la qualité d’images sont utilisées comme fonctions de coûts (intégrées dans le critère global) et ce afin d’étudier leur potentiel pour améliorer les performances de restauration. Dans le contexte où l’image d’intérêt est corrompue par d’autres images interférentes, sa restauration nécessite le recours aux techniques de séparation aveugle de sources. Pour cela, une étude comparative de certaines techniques de séparation basées sur la propriété de décorrélation au second ordre et la parcimonie est réalisée
This thesis concerns the blind restoration of images (formulated as an ill-posed and illconditioned inverse problem), considering a SIMO system. Thus, a blind system identification technique in which the order of the channel is unknown (overestimated) is introduced. Firstly, a simplified version at reduced cost SCR of the cross relation (CR) method is introduced. Secondly, a robust version R-SCR based on the search for a sparse solution minimizing the CR cost function is proposed. Image restoration is then achieved by a new approach (inspired from 1D signal decoding techniques and extended here to the case of 2D images) based on an efficient tree search (Stack algorithm). Several improvements to the ‘Stack’ method have been introduced in order to reduce its complexity and to improve the restoration quality when the images are noisy. This is done using a regularization technique and an all-at-once optimization approach based on the gradient descent which refines the estimated image and improves the algorithm’s convergence towards the optimal solution. Then, image quality measurements are used as cost functions (integrated in the global criterion), in order to study their potential for improving restoration performance. In the context where the image of interest is corrupted by other interfering images, its restoration requires the use of blind sources separation techniques. In this sense, a comparative study of some separation techniques based on the property of second-order decorrelation and sparsity is performed

Submitted by Diogo Barreiros (diogo.barreiros@ufba.br) on 2017-03-10T14:39:59Z No. of bitstreams: 1 tese_alexandre_versao_final_bd.pdf: 3780030 bytes, checksum: 765f095f9626a12f3b43a6bf9fdb97f3 (MD5)
Approved for entry into archive by Vanessa Reis (vanessa.jamile@ufba.br) on 2017-03-10T14:52:25Z (GMT) No. of bitstreams: 1 tese_alexandre_versao_final_bd.pdf: 3780030 bytes, checksum: 765f095f9626a12f3b43a6bf9fdb97f3 (MD5)
Made available in DSpace on 2017-03-10T14:52:25Z (GMT). No. of bitstreams: 1 tese_alexandre_versao_final_bd.pdf: 3780030 bytes, checksum: 765f095f9626a12f3b43a6bf9fdb97f3 (MD5)
The automatic person re-identification (re-id) problem resides in matching an unknown person image to a database of previously labeled images of people. Among several issues to cope with this research field, person re-id has to deal with person appearance and environment variations. As such, discriminative features to represent a person identity must be robust regardless those variations. Comparison among two image features is commonly accomplished by distance metrics. Although features and distance metrics can be handcrafted or trainable, the latter type has demonstrated more potential to breakthroughs in achieving state-of-the-art performance over public data sets. A recent paradigm that allows to work with trainable features is deep learning, which aims at learning features directly from raw image data. Although deep learning has recently achieved significant improvements in person re-identification, found on some few recent works, there is still room for learning strategies, which can be exploited to increase the current state-of-the-art performance. In this work a novel deep learning strategy is proposed, called here as coarse-to-fine learning (CFL), as well as a novel type of feature, called convolutional covariance features (CCF), for person re-identification. CFL is based on the human learning process. The core of CFL is a framework conceived to perform a cascade network training, learning person image features from generic-to-specific concepts about a person. Each network is comprised of a convolutional neural network (CNN) and a deep belief network denoising autoenconder (DBN-DAE). The CNN is responsible to learn local features, while the DBN-DAE learns global features, robust to illumination changing, certain image deformations, horizontal mirroring and image blurring. After extracting the convolutional features via CFL, those ones are then wrapped in covariance matrices, composing the CCF. CCF and flat features were combined to improve the performance of person re-identification in comparison with component features. The performance of the proposed framework was assessed comparatively against 18 state-of-the-art methods by using public data sets (VIPeR, i-LIDS, CUHK01 and CUHK03), cumulative matching characteristic curves and top ranking references. After a thorough analysis, our proposed framework demonstrated a superior performance.

Darknet markets are online services behind Tor where cybercriminals trade illegal goods and stolen datasets. In recent years, security analysts and law enforcement start to investigate the darknet markets to study the cybercriminal networks and predict future incidents. However, vendors in these markets often create multiple accounts (i.e., Sybils), making it challenging to infer the relationships between cybercriminals and identify coordinated crimes. In this thesis, we present a novel approach to link the multiple accounts of the same darknet vendors through photo analytics. The core idea is that darknet vendors often have to take their own product photos to prove the possession of the illegal goods, which can reveal their distinct photography styles. To fingerprint vendors, we construct a series deep neural networks to model the photography styles. We apply transfer learning to the model training, which allows us to accurately fingerprint vendors with a limited number of photos. We evaluate the system using real-world datasets from 3 large darknet markets (7,641 vendors and 197,682 product photos). A ground-truth evaluation shows that the system achieves an accuracy of 97.5%, outperforming existing stylometry-based methods in both accuracy and coverage. In addition, our system identifies previously unknown Sybil accounts within the same markets (23) and across different markets (715 pairs). Further case studies reveal new insights into the coordinated Sybil activities such as price manipulation, buyer scam, and product stocking and reselling.
Master of Science
Taking advantage of the high anonymity of darknet, cybercriminals have set up underground trading websites such as darknet markets for trading illegal goods. To understand the relationships between cybercriminals and identify coordinated activities, it is necessary to identify the multiple accounts hold by the same vendor. Apart from manual investigation, previous studies have proposed methods for linking multiple accounts through analyzing the writing styles hidden in the users' online posts, which face key challenges in similar tasks on darknet markets. In this thesis, we propose a novel approach to link multiple identities within the same darknet market or across different markets by analyzing the product photos. We develop a system where a series of deep neural networks (DNNs) are used with transfer learning to extract distinct features from a vendor's photos automatically. Using real-world datasets from darknet markets, we evaluate the proposed system which shows clear advantages over the writing style based system. Further analysis of the results reported by the proposed system reveal new insights into coordinated activities such as price manipulation, buyer scam and product stocking and reselling for those vendors who hold multiple accounts.

Cette thèse s'inscrit dans le contexte de la vidéo surveillance et s'intéresse à la ré-identification de personnes dans un réseau de caméras à champs disjoints. La ré-identification consiste à déterminer si une personne quitte le champ d'une caméra et réapparait dans une autre. Elle est particulièrement difficile car l'apparence de la personne change de manière significative à cause de différents facteurs. Nous proposons d'exploiter la complémentarité de l'apparence de la personne et son style de mouvement pour la décrire d'une manière appropriée aux facteurs de complexité. C'est une nouvelle approche car la ré-identification a été traitée par des approches d'apparence. Les contributions majeures proposées concernent: la description de la personne et l'appariement des primitives. Nous étudions deux scénarios de ré-identification : simple et complexe. Dans le scénario simple, nous étudions la faisabilité de deux approches : approche biométrique basée sur la démarche et approche d'apparence fondée sur des points d'intérêt (PI) spatiaux et des primitives de couleur. Dans le scénario complexe, nous proposons de fusionner des primitives d'apparence et de mouvement. Nous décrivons le mouvement par des Pis spatio-temporels et l'apparence par des PIs spatiaux. Pour l'appariement, nous utilisons la représentation parcimonieuse comme méthode d'appariement local entre les PIs. Le schéma de fusion est fondé sur le calcul de la somme pondérée des votes des PIs et ensuite l'application de la règle de vote majoritaire. Nous proposons également une analyse d'erreurs permettant d'identifier les sources d'erreurs de notre système pour dégager les pistes d'amélioration les plus prometteuses
This thesis focuses on the problem of hu man re-identification through a network of cameras with non overlapping fields of view. Human re-identification is defined as the task of determining if a persan leaving the field of one camera reappears in another. It is particularly difficult because of persons' significant appearance change within different cameras vision fields due to various factors. In this work, we propose to exploit the complementarity of the person's appearance and style of movement that leads to a description that is more robust with respect to various complexity factors. This is a new approach for the re-identification problem that is usually treated by appearance methods only. The major contributions proposed in this work include: person's description and features matching. First we study the re-identification problem and classify it into two scenarios: simple and complex. In the simple scenario, we study the feasibility of two approaches: a biometric approach based on gait and an appearance approach based on spatial Interest Points (IPs) and color features. In the complex scenario, we propose to exploit a fusion strategy of two complementary features provided by appearance and motion descriptions. We describe motion using spatiotemporal IPs, and use the spatial IPs for describing the appearance. For feature matching, we use sparse representation as a local matching method between IPs. The fusion strategy is based on the weighted sum of matched IPs votes and then applying the rule of majority vote. Moreover, we have carried out an error analysis to identify the sources of errors in our proposed system to identify the most promising areas for improvement

Cette thèse CIFRE est effectuée dans un contexte industriel et présente un framework complet pour la détection, le suivi mono-caméra et de la ré-identification de personnes dans le contexte multi-caméras. Les performances élevés et le traitement en temps réel sont les deux contraintes critiques ayant guidé ce travail. La détection de personnes vise à localiser/délimiter les gens dans les séquences vidéo. Le détecteur proposé est basé sur une cascade de classifieurs de type LogitBoost appliqué sur des descripteurs de covariances. Une approche existante a fortement été optimisée, la rendant applicable en temps réel et fournissant de meilleures performances. La méthode d'optimisation est généralisable à d'autres types de détecteurs d'objets. Le suivi mono-caméra vise à fournir un ensemble d'images de chaque personne observée par chaque caméra afin d'extraire sa signature visuelle, ainsi qu'à fournir certaines informations du monde réel pour l'amélioration de la ré-identification. Ceci est réalisé par le suivi de points SIFT à l'aide d'une filtre à particules, ainsi qu'une méthode d'association de données qui infère le suivi des objets et qui gère la majorité des cas de figures possible, notamment les occultations. Enfin, la ré-identification de personnes est réalisée avec une approche basée sur l'apparence globale en améliorant grandement une approche existante, obtenant de meilleures performances tout en étabt applicable en temps réel. Une partie "conscience du contexte" est introduite afin de gérer le changement d'orientation des personnes, améliorant les performances dans le cas d'applications réelles.

碩士
國立臺灣大學
光電工程學研究所
105
Artificial Intelligence (A.I.) and technologies of machine learning had made significant success in recent years. Many of things that are known as “human-operating-only” can be done well by A.I. Although self-driving cars are not widespread now, it’s can be anticipated that self-driving will become more popular as the technology is improved in the future. Vehicle-to-vehicle (V2V) communication is one of the important parts of self-driving. The positions and velocities of nearby cars can be detected by V2V communication, and the information can be further used to prevent collisions. Visible light communication (VLC) has high potential in this field. One of the reasons is that most of cars are using light-emitting-diodes (LEDs) for illumination and LEDs are easily modulated and have high efficiency, which is a perfect transmitter for VLC. The other reason is that visible light can be blocked easily and has high directivity. So that only nearby signals generated in proper orientations are detected, which is an important benefit in V2V communication. In this thesis, an Arduino board was used to modulate arrays of LEDs, and use smart phone camera as the receiver to build a VLC system. This device is used to simulate the actual situation of VLC between the tail-light of a car and a dash cam. In this thesis, rolling-shutter effect and image processing are used to identify difference between normal light source and the modulated tail-light of car. After this, boundary detection is used to decode message and estimate the distance between cars. For image processing, adaptive the histogram equalization and the two-dimensional convolution are used to strengthen the ability of the system to identifying modulated tail-light. On-off keying is used for modulation, and run-length limited (RLL) coding prevents too many continuous zeros affecting illumination and decoding signal. In the end, limited by the frame rate of camera, this device can only send 16 bits per second, which is designed to transmit velocity information of car.

Η αυτόματη ταυτοποίηση εικόνων δακτυλικών αποτυπωμάτων αποτελεί ένα δύσκολο και πολυδιάστατο πρόβλημα, το οποίο έχει απασχολήσει πλήθος ερευνητών και για το οποίο έχει αναπτυχθεί μεγάλος αριθμός τεχνικών. Η δυσκολία του προβλήματος έγκειται στο γεγονός ότι οι εικόνες των αποτυπωμάτων είναι σε μεγάλο ποσοστό αλλοιωμένες ή ακόμα σε κάποιες περιπτώσεις δεν είναι διαθέσιμη η πλήρης εικόνα του αποτυπώματος, αλλά μόνο ένα μέρος αυτής. Στη συγκεκριμένη διατριβή, προτείνονται δύο μέθοδοι αυτόματης ταυτοποίησης δακτυλικών αποτυπωμάτων: α) η μέθοδος ταυτοποίησης δακτυλικών αποτυπωμάτων με χρήση τεχνικών ευθυγράμμισης και β) μέθοδος ταυτοποίησης δακτυλικών αποτυπωμάτων από το συνδυασμό του Δικτύου Αυτό-Οργανούμενων Δικτύων του Kohonen και του ορισμού των σημείων μικρολεπτομερειών των αποτυπωμάτων ως νευρώνων του δικτύου. Επιπλέον, ιδιαίτερη βαρύτητα δόθηκε στην προεπεξεργασία των εικόνων των δακτυλικών αποτυπωμάτων βάσει της ανάπτυξης και εφαρμογής κατάλληλων τεχνικών επεξεργασίας εικόνων προκειμένου να προκύψει βελτίωση της ποιότητας της εικόνας του δακτυλικού αποτυπώματος και να εξαχθούν οι μικρολεπτομέρειες που χρησιμοποιούνται για την ταύτιση των δακτυλικών αποτυπωμάτων. Στο πλαίσιο της παρούσας διατριβής, χρησιμοποιήθηκαν δεδομένα δακτυλικών αποτυπωμάτων από τις βάσεις VeriFingerSample_DB της Neurotechnology και η DB3 του διαγωνισμού δακτυλικών αποτυπωμάτων FVC2004. Για την ποσοτική αποτίμηση της απόδοσης των προτεινόμενων μεθόδων χρησιμοποιήθηκε το κριτήριο της Αναλογίας Ίσου Σφάλματος (EqualErrorRate – EER). Σύμφωνα με το κριτήριο αυτό, η μέθοδος ταυτοποίησης δακτυλικών αποτυπωμάτων βάσει του Δικτύου Αυτό-Οργανούμενων Δικτύων παρείχε καλύτερα αποτελέσματα σε σύγκριση με οποιαδήποτε μέθοδο ευθυγράμμισης που εφαρμόστηκε.
Automatic Fingerprint Identification is a difficult and multidimensional problem. For this reason, the number of papers and techniques regarding this issue is numerous. The hardness of the problem lies with the fact that there is a large percentage of corrupted and partial fingerprint images. Throughout this Thesis, two methods were proposed for the Automatic Fingerprint Identification: a) the Automatic Fingerprint Identification based on registration techniques and b) the Automatic Fingerprint Identification based on the theory of Self Organizing Maps of Kohonen, setting the minutiae of the fingerprint images as input neurons of the Map. Furthermore, an important step prior to the application of the proposed automatic fingerprint identification methods is the pre-processing of these images by the development and implementation of a series of image processing techniques in order to enhance the image quality and to extract the minutiae which are then used for the fingerprint identification. In this Thesis, a substantial number of fingerprint images were used from the database VeriFingerSample_DB kai from the database DB3 of the competition FVC2004. The quantitative evaluation of both proposed automatic fingerprint identification methods were based on the Equal Error Rate (EER) criterion. According to this, the Automatic Fingerprint Identification based on the Self Organizing Maps outperformed against any other method based on registration techniques.

碩士
輔仁大學
電機工程學系碩士班
107
Neural network methods of person re-identification are mostly based on the inception structure, which leads to the huge amount of model parameters and the problem of gradient vanishing. DenseNet solves the problem of gradient vanishing by proposing dense block, but dense block has a high demand for GPU memory, making it difficult to use. For the multi-task neural network methods, research on feature fusion is lack of experiments and detailed discussion. The pyramid depth parameters proposed in this paper can reduce the GPU memory required for the dense block-based network model and maintain the accuracy of the model. This paper also proposes the parameter efficiency type dense block that further reduces the number of parameters. Because of these two designs, the dense block can be used for complex network structures. This paper collects a variety of fusion units, proposes three fusion structures. We find the most effective feature fusion subnet by analyzing the feature fusion subnets composed of different fusion units and fusion structures. Through the above improvement, this paper implemented a neural network method of person re-identification based on body parts-BPMDN. The feature extraction subnet of BPMDN can be implemented because of pyramid depth parameters and parameter efficiency type dense block. BPMDN saves about half the amount of parameters and has higher accuracy than inception-based methods, and shows excellent generalization ability compared with state-of-the-art.

碩士
國立臺灣師範大學
體育學系
101
The purpose of this study was to reveal the relations among team image, team identification, loyalty and team re-support intension in fans of EDA Rhinos. Five hundred questionnaires were sent during March 18 to 27, 2013. Descriptive statistics, multiple stepwise regression and logistic regression were used for data analysis. The results were stated as follows: 1. The main group of fans in EDA Rhinos was college graduated student less than 28 years old, TV as their view massage, and their monthly salary are below NT 10,000 dollars. 2. The point of team image, team identification and loyalty were brightly higher than average in this study. 3. Team identification and loyalty could be well predicted by team image. 4. Team identification was the best predictor for loyalty. 5. Team image and loyalty affect re-support intension positively, and team image and loyalty are main predictors. According to the results, EDA Rhinos should strengthen the characteristic management of home court, as the emotion connection would positively affect team image and loyalty, there will be a beneficial impact on whole management.

Thesis (M.S.)--State University of New York at Buffalo, 2005.
Title from PDF title page (viewed on Apr. 13, 2006) Available through UMI ProQuest Digital Dissertations. Thesis adviser: Jacobi, Robert D. Includes bibliographical references.