The user perspective of image search remains poorly understood. the purpose of this study is to identify and investigate the key issues relevant to a user’s interaction with images and the user’s approach to image search. a deeper understanding of these issues will serve to inform the design of image retrieval systems and in turn better serve the user. Previous research explores areas of information seeking behavior, representation in information science, query formulation, and image retrieval. the theoretical framework for this study includes an articulation of image search scenarios as adapted from Yoon and O’Connor’s taxonomy of image query types, Copeland’s Engineering Design Approach for rigorous qualitative research, and Anderson’s Functional Ontology Construction Model for building robust models of human behavior. a series of semi-structured interviews were conducted with expert-level image users. Interviewees discussed their motivations for image search, types of image searches they pursue, and varied approaches to image search, as well as how they decide that an information need has been met and which factors influence their experience of search. a content analysis revealed themes repeated across responses, including a collection of 23 emergent concepts and 6 emergent categories. a functional analysis revealed further insight into these themes. Results from both analyses may be used as a framework for future exploration of this topic. Implications are discussed and future research directions are indicated. Among possibilities for future research are investigations into collaborative search and ubiquitous image search.

Image segmentation is a fundamental part in low level computer vision processing. It has an essential influence on the subsequent higher level visual scene interpretation for a wide range of applications. Unsupervised image segmentation is an ill-defined problem and thus cannot be optimally solved in general. Several novel unsupervised multispectral image segmentation methods based on the underlaying random field texture models (GMRF, 2D/3D CAR) were developed. These segmenters use efficient data representations that allow an analytical solutions and thus the segmentation algorithm is much faster in comparison to methods based on MCMC. All segmenters were extensively compared with the alternative state- of-the-art segmenters with very good results. The MW3AR segmenter scored as one of the best available. The cluster validation problem was solved by a modified EM algorithm. Two multiple resolution segmenters were designed as a combination of a set of single segmenters. To tackle a realistic variable lighting in images, the illumination invariant features were derived and the illumination invariant segmenter was developed. For the proper evaluation of segmentation results and ranking of algorithms, a unique web-based texture segmentation benchmark was proposed and implemented. It was used for comprehensive...

Ces dernières décennies ont vu émerger de nombreuses applications utilisant la technologie 3D, telles que les écrans auto-stéréoscopiques, les écrans de télévisions 3D ou encore la visio-conférence stéréoscopique. Ces applications requièrent des techniques adaptées afin que leur flux de données soit compressé efficacement. En particulier, dans le cas des images stéréoscopiques, ces dernières étant composées de deux vues de la même scène, elles nécessitent à ce titre deux fois plus d’informations à transmettre ou à stocker que dans le cas des images 2D traditionnelles. Nos travaux se situent dans le cadre de la compression des images stéréoscopiques. Plus précisément, ils concernent l’amélioration de l’estimation de la carte de disparité dans le but d’obtenir un meilleur compromis entre le débit binaire nécessaire au codage de la carte de disparité et la qualité de l’image prédite.Généralement, la carte de disparité est estimée en minimisant la distorsion de l’image prédite. Cette minimisation peut être sujette à une contrainte de lissage. L’idée étant qu’une carte de disparité plus lisse nécessitera un débit binaire moindre en supposant que les mêmes vecteurs de disparités seront sélectionnés plus souvent. Néanmoins cette contrainte de lissage ne permet pas toujours de diminuer le coût binaire de la carte. Le lissage peut entraîner par ailleurs une augmentation notable de la distorsion de l’image prédite. Dans le premier chapitre de la thèse, nous présentons un algorithme d’estimation de carte de disparité minimisant une métrique conjointe entropie-distorsion. Le coût binaire finale de la carte de disparité est estimée à chaque étape de l’algorithme et est intégré dans le calcul de la métrique. La distorsion globale de la carte de disparité est aussi mise à jour au fur et à mesure du traitement de l’image. Par ailleurs, cette algorithme repose sur la construction séquentiel d’un arbre dont on ne garde qu’un nombre défini de branches à chaque profondeur de l’arbre. Ainsi, l’algorithme développé apporte une solution sous-optimale en minimisant le coût binaire de la carte de disparité tout en assurant une bonne qualité de l’image prédite. Le chapitre deux étend l’algorithme précédent au cas des images non rectifiées. Dans le troisième chapitre, nous nous intéressons au fait de trouver une solution au problème d’optimisation du compromis débit-distorsion en réduisant la complexité numérique par rapport à l’algorithme précédent. De ce fait, nous avons développé le R-algorithme qui se base sur une solution initiale de Référence (celle minimisant la distorsion de l’image prédite) et la modifie successivement tant qu’une amélioration est constatée en termes de compromis débit-distorsion. Le quatrième chapitre s’intéresse toujours au fait d’accroître les performances de l’algorithme développé tout en réduisant le coût en complexité numérique et donc en temps de traitement. Nous proposons deux approches afin de tirer profit d’un grand espace de recherche sans avoir à tester pour chaque bloc à apparier l’ensemble des disparités qui composent cet espace de recherche. En effet, un espace de recherche plus grand permet plus de choix de disparités et donc potentiellement une meilleur reconstruction de l’image prédite. En contrepartie, il se peut que le coût binaire de la carte de disparité augmente si l’ensemble des disparités sélectionnées constituent un ensemble plus divers qu’auparavant. Les deux approches proposées permettent de restreindre l’espace de recherche à un ensemble composées de certaines disparités permettant de minimiser la distorsion de l’image prédite pour un débit donné. Le dernier chapitre de la thèse s’intéresse à l’utilisation des blocs de taille variable pour la compression des images stéréoscopiques. / During the last decades, a wide range of applications using stereoscopic technology has emerged still offering an increased immersion to the users such as video games with autostereoscopic displays, 3D-TV or stereovisio-conferencing. The raise of these applications requires fast processing and efficient compression techniques. In particular, stereoscopic images require twice the amount of information needed to transmit or store them in comparison with 2D images as they are composed of two views of the same scene. The contributions of our work are in the field of stereoscopic image compression and more precisely, we get interested in the improvement of the disparity map estimation. Generally, disparities are selected by minimizing a distortion metric which is sometimes subjected to a smoothness constraint, assuming that a smooth disparity map needs a smaller bitrate to be encoded. But a smoother disparity map does not always reduce significantly the bitrate needed to encode it but can increase the distortion of the predicted view. Therefore, the first algorithm we have proposed minimizes a joint entropy-distortion metric to select the disparities. At each step of the algorithm, the bitrate of the final disparity map is estimated and included in the metric to minimize. Moreover, this algorithm relies on a tree where a fixed number of paths are extended at each depth of the tree, ensuring good rate-distortion performance. In the second part of the work, we have proposed a sub-optimal solution with a smaller computational complexity by considering an initial solution -the one minimizing the distortion of the predicted view- which is successively modified as long as an improvement is observed in terms of rate-distortion. Then, we have studied how to take advantages of large search areas in which the disparities are selected as one can easily supposed that enlarging the search area will increase the distortion performance as there will be more choices of disparities. In the other hand, the larger is the range of the selected disparities, the higher is supposed to be the cost of the disparity map in terms of bitrate. We have proposed two approaches allowing to take advantage of a large search area by selecting only sets of disparities belonging to it enabling to achieve a given bitrate while minimizing the distortion of the predicted image. The last part of the work concerns variable block sizes which undeniably allows to improve the bitrate-distortion performance as the block size suits to the image features. We have thus proposed a novel algorithm which jointly estimates and optimizes the disparity and the block length maps.

Bibliography: pages 174-182. / This dissertation investigates the machine vision grading of flue-cured Virginia tobacco by means of digital processing of tobacco leaf images. With reference to international grading standards and to modem image processing techniques, two classifiers are designed. The colour classifier uses seven features extracted from each leaf image to grade the leaf into one of five official colour classes. It does this with an expected correct classification rate of 93.5%. The plant position classifier identifies the position on the stalk from which a leaf was reaped, using ten size and shape features to classify the leaf into one of six plant position categories. It has a correct classification rate of 70%. Average colours for each colour class and archetypal shapes for each plant position category are derived from the digital leaf data. These should be of value to tobacco graders as objective representations of typical leaves within each class.

Tensors are of great interest to many applications in engineering and in medical imaging, but a proper analysis and visualization remains challenging. It already has been shown that, by employing the metaphor of a fabric structure, tensor data can be visualized precisely on surfaces where the two eigendirections in the plane are illustrated as thread-like structures. This leads to a continuous visualization of most salient features of the tensor data set. We introduce a novel approach to compute such a visualization from tensor field data that is motivated by image-space line integral convolution (LIC). Although our approach can be applied to arbitrary, non-selfintersecting surfaces, the main focus lies on special surfaces following important features, such as surfaces aligned to the neural pathways in the human brain. By adding a postprocessing step, we are able to enhance the visual quality of the of the results, which improves perception of the major patterns.

La découverte d'archives photographiques militaires de la première guerre du Golfe constitue le point de départ de cette recherche. A partir de l'analyse de ces documents, il s'est agi d'étudier les différents régimes de visibilité mis à l'oeuvre dans les représentations actuelles des conflits, et de comprendre jusqu'à quel point la guerre donne à voir d'elle-même, à travers les images qu'elle produit. Considérée comme un ensemble de dispositifs de visualisation et de représentation, la guerre produit un visuel. Dès lors, qu'est-ce que la guerre dévoile dans ce qu'elle donne à voir et ce qu'elle ne montre pas ?La guerre doit être pensée en lien avec notre regard. Depuis la guerre du Golfe, conflit qui a marqué une rupture dans la façon de penser et de représenter la guerre, le monde est devenu un écran géant sur lequel se succèdent, et de façon de plus en plus rapide, un flux et un flot d'images ininterrompus. L'image de guerre, telle qu'elle était pensée, montrée, imaginée jusque-là, n'allait plus jamais être la même. Le conflit de 1990 dans le Golfe n'a rien montré. "On n'a rien vu", sont les mots les plus souvent utilisés pour décrire ce conflit, dont la représentation était totalement absente :elle était hors-champ ou plus précisément, une représentation nous était donnée, mais la réalité de la guerre était maintenue hors champ. Le concept de dé-subjectivation évoqué soulève bien évidemment des questions morales décisives en particulier, mais pas seulement, la question de la responsabilité, responsabilité des militaires, des politiques, des journalistes, mais également des artistes lorsqu'ils travaillent avec ce matériau de documentation. / Doctorat en Art et Sciences de l'Art / info:eu-repo/semantics/nonPublished

<p>Image compression is widely applied to medical imaging, remote sensing applications, biomedical diagnosis, multimedia applications and so on [1]-[4]. In many cases, considering the factor of image quality, we use a lossless compression method to compress the image.</p> <p>In this thesis work, we propose bit-error aware lossless compression algorithms for color image compression subject to bit-error rate during transmission. Each of our proposed algorithms includes three stages. The first stage is to convert the RGB images to YCrCb images, and the second stage predicts the transformed images to generate the residue sequences. Optimization algorithms are used to search the best combination of the image conversion and prediction. At the last stage, <a>the generated residue sequences are encoded by several residue coding algorithms, which are 2-D and 1-D bi-level block coding, interval Huffman coding and standard Huffman coding algorithms</a>. Key parameters, such as color transformation information, predictor parameters and residue coding parameters, are protected by using (7,4) Hamming code during image transmission, </p> <p><a>The compression ratio (CR) and peak signal to noise ratio (PSNR)</a> are two significant performance indicators which are used to evaluate the experimental results. According to the experimental results, the 2-D bi-level block coding algorithm is verified as the best coding method.</p>

Indiana University-Purdue University Indianapolis (IUPUI) / A pulsatile flow loop can be suitable for measurement of in vitro blood pressure. The pressure data collected from such a system can be used for evaluating stenosis in human arteries, a condition in which the arterial lumen size is reduced. The objective of this work is to develop an experimental system to simulate blood flow in the human arterial system. This system will measure the in vitro hemodynamics using 3-D prints of vessels extracted from patient CT images. Images are segmented and processed to produce 3-D prints of vessel geometry, which are mounted in the loop. Control of flow and pressure is made possible by the use of components such as a pulsatile heart pump, resistance, and compliance elements. Output data is evaluated by comparison with CFD and invasive measurement. The system is capable of measurement of the pressures such as proximal, Pa, and distal, Pd, pressures to evaluate in vivo conditions and to assess the severity of stenosis. This is determined by use of parameters such as fractional flow reserve (FFR=Pd/Pa) or trans-stenotic pressure gradient (TSPG=Pa-Pd). This can be done on a non-invasive, patient specific basis, to avoid the risk and high cost of invasive measurement. In its operation, the preliminary measurement of blood pressures demonstrates agreement with the invasive measurement as well as the CFD results. These preliminary results are encouraging and can be improved upon by continuing development of the experimental system. A working pulsatile loop has been reached, an initial step taken for continued development. This loop is capable of measuring the flow and pressure from in a 3-D printed artery. Future works will include more life-like material for the artery prints, as well as cadaver vessels.

As technology has advanced, data has become more and more important. The more breakthroughs are achieved, the more data is needed to support them. As a result, more storage is required in the system's memory. Compression is therefore required. Before it can be stored, the data must be compressed. To ensure that information is not lost, efficient compression is necessary. This also makes sure that there is no redundancy in the data that is being kept and stored. Compressive sensing has emerged as a new field of compression thanks to developments in sparse optimization. Rather than relying just on compression and sensing formulations, the theory blends the two. The objective of this thesis is to analyze the concept of compressive sensing and to study several reconstruction algorithms. Additionally, a few of the algorithms were put into practice. This thesis also included a model of the ECG, which is vital in determining the health of the heart. For the most part, the ECG is utilized to diagnose heart illness, and a modified synthetic ECG can be used to mimic some of these arrhythmias.

Submitted to the Faculty of Education in fulfillment of the requirements for the Master of Education in the Department of Educational Psychology and Special Education at the University of Zululand, South Africa, 2000. / This study focuses on how drum majoretting may influence the person image of the adolescent girl. An attempt was made to determine or investigate the influence of relationship formation on the person image and whether drum majoretting influences relationship formation. The researcher also investigated the influence of drum majoretting on the person image. The fact that different concepts are used to describe this totality (person image) can easily lead to confusion. Consequently it was necessary to review the nature and influence of the most important areas that are related to the person image of the adolescent drum majorette: ► relationship formation ► physical development cognitive development emotional development ► social development. The literature study discloses that the drum majorette is involved in relationship formation with self, peers, trainer/educator and parents. Analysis of the data indicates that the adolescent whose relationships with significant others were favourable, positive and rewarding would have developed a positive self-identity, a positive self-concept and an enhanced feeling of self. Unfavourable experiences in the relationship formation are likely to limit the drum majorette's development and an impaired self-concept, negative self-identity would have been formed. Relationships in the life-world of the drum majorette may have a significant effect on the person image. Drum majoretting as a sport provides a range of experiences through which the adolescent may be socialised. For the purpose of the empirical investigation a self-structured questionnaire for drum majorettes were utilised. The questionnaires that were completed by drum majorettes were analyzed and thereafter the data that was obtained was processed and analyzed. The study has confirmed that the person image of the adolescent girl is enhanced through drum majoretting. The empirical investigation proved that all four categories of relationship i.e. self, peers, trainer/educator and parents are important to the adolescent drum majorette in order for her to self-actualize. Findings also show that the physical, cognitive, emotional and social development of the drum majorette is enhanced. In conclusion a summary .was presented and based on the findings of this study, the following are some of the recommendations that were made; ► The Department of Education should consider drum majoretting as a compulsory extra-curricular code of sport for girls. ► The school (principal) should develop and enforce a policy which will provide a set of guidelines to organize the code of drum majoretting through which the person image can be enhanced. ► A proactive empowerment programme should be established where parents can be addressed on the role of drum majoretting. Parent-child relationships are fundamental to the pupil's development of the person image. The parent-child contact will bring the educator into focus. A good parent-educator relationship is vital for the enhancement of the person image of the adolescent drum majorette. ► A strategic plan should be formulated for drum majoretting with the cooperation of parents, educators and the Drum Majorette Association of South Africa.