• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 14
  • 12
  • 10
  • 8
  • 4
  • 1
  • 1
  • 1
  • 1
  • 1
  • Tagged with
  • 57
  • 10
  • 9
  • 7
  • 7
  • 5
  • 5
  • 5
  • 4
  • 4
  • 4
  • 4
  • 4
  • 4
  • 4
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
41

Detekce význačných bodů v obraze / Interest Point Detection in Images

Duda, Tomáš Unknown Date (has links)
This master's thesis deals with the interest point detection in images. The main goal is to create an application for making panoramic photos, which is based on this detection. The application uses the SIFT detector for finding keypoints of the image. Afterwards the geometrical consistence of image points using the RANSAC algorithm is finds out and images are transform into panorama in accordance with this geometrical consistence. Finally techniques for blending of transition between photos are used.
42

Kolektivizace zemědělství a její promítnutí do krajiny. Případová studie katastrálního území Široký Důl / Collectivization of agriculture and its projection to the landscape. Case study of cadastral area Široký Důl

Kocmanová, Pavlína January 2010 (has links)
Diplomová práce Kolektivizace zemědělství a její promítnutí do krajiny ABSTRACT I searched changes in the cultural landscape structure in the relation with social processes in the time period 1937-2010 in my thesis. The research refers to the second half of 20th century, or more precisely, to the era of collectivization of Czechoslovak agriculture. In consequence of collectivization, the landscape was ecologically devastated and overall the nature of Czech countryside changed. Analysis of the landscape structure has form of case study of cadastral area of Široký Důl. I interpreted landscape changes from historical aerial photos and then analyzed the changes by geographical information systems. The results of those analyses are indicators (landscape metrics), by which I interpreted conditions of landscape in particular time. I confronted the changes in the landscape structure with social processes, which are described in theoretical part of thesis according to background research. Result of my thesis is general story of particular landscape, whose structure changed fundamentally in consequence of collectivization of agriculture. Key words: landscape structure, collectivization, landscape metrics, GIS, cultural landscape, aerial photos, Široký Důl
43

"En skål för andras lidande" : Bildanalys av vinnarbilderna i World Press Photo of the Year mellan 1999-2008

Rodling, Emma, Ekman, Maria January 2010 (has links)
<p>En undersökning av vinnarbilderna i World Press Photo of the Year mellan 1999-2008. World Press Photo är världens största fototävling och är prestigefylld att vinna för pressfotografer. Syftet var att undersöka vad de senaste årens bildjournalistik förmedlar till sina betraktare genom nyhetsbilderna som vunnit World Press Photo of the Year under 2000-talet. Syftet har förgrenat sig i frågeställningarna:</p><p>Vilka teman har vinnarbilderna i World Press Photo of the Year mellan 1999-2008?</p><p>Vad förmedlar innehållet i bilderna?</p><p>Vad finns det för likheter och skillnader mellan bilderna?</p><p>Studien är en kvalitativ undersökning där vi gjort bildanalyser av vinnarbilderna i World Press Photo of the Year mellan 1999-2008. Vi har använt oss av semiotisk och retorisk bildanalys. Resultaten visar att vinnarbilderna som representerar det bästa inom bildjournalistiken visar ett begränsat antal teman, krig, sociala problem och fattigdom samt naturkatastrofer. De flesta bilderna är från Asien och framförallt Mellanöstern. Bilderna är negativt laddade och människorna på bilderna visar oftast sorg och lidande.</p>
44

"En skål för andras lidande" : Bildanalys av vinnarbilderna i World Press Photo of the Year mellan 1999-2008

Rodling, Emma, Ekman, Maria January 2010 (has links)
En undersökning av vinnarbilderna i World Press Photo of the Year mellan 1999-2008. World Press Photo är världens största fototävling och är prestigefylld att vinna för pressfotografer. Syftet var att undersöka vad de senaste årens bildjournalistik förmedlar till sina betraktare genom nyhetsbilderna som vunnit World Press Photo of the Year under 2000-talet. Syftet har förgrenat sig i frågeställningarna: Vilka teman har vinnarbilderna i World Press Photo of the Year mellan 1999-2008? Vad förmedlar innehållet i bilderna? Vad finns det för likheter och skillnader mellan bilderna? Studien är en kvalitativ undersökning där vi gjort bildanalyser av vinnarbilderna i World Press Photo of the Year mellan 1999-2008. Vi har använt oss av semiotisk och retorisk bildanalys. Resultaten visar att vinnarbilderna som representerar det bästa inom bildjournalistiken visar ett begränsat antal teman, krig, sociala problem och fattigdom samt naturkatastrofer. De flesta bilderna är från Asien och framförallt Mellanöstern. Bilderna är negativt laddade och människorna på bilderna visar oftast sorg och lidande.
45

Analyse régionale de la structure et de la dynamique biogéomorphologiques des rivières en tresses du bassin du Rhône / Regional analysis of the biogeomorphological structure and dynamic of braided rivers in the Rhône basin district

Belletti, Barbara 19 December 2012 (has links)
Le paysage des rivières en tresses est constitué d’une riche mosaïque d’habitats diversifiés, déterminée par un ensemble de facteurs biophysiques qui interagissent à différentes échelles spatiales et temporelles.A partir d’une sélection de 53 tronçons en tresses, nous avons réalisé une analyse comparative à l’échelle régionale du bassin du Rhône et en fonction de plusieurs échelles temporelles, en mobilisant les techniques de télédétection.Premièrement, nous avons étudié la variabilité de la structure des habitats aquatiques et du patron des chenaux. Nous montrons que le patron de tressage et des habitats aquatiques n’est pas uniquement lié au débit, mais aussi aux conditions locales liées aux apports sédimentaires et à la présence de la nappe phréatique.Deuxièmement, nous avons abordé une approche diachronique afin de comprendre l’évolution à moyen terme du paysage fluvial de ces tronçons, depuis la moitié du siècle dernier. Les résultats montrent que toutes les tresses du bassin rhodanien ne sont pas déliquescentes et que certains tronçons sont encore très actifs. Cela dépend d’un ensemble de facteurs : la position géographique du tronçon, son évolution à court terme liée aux épisodes de crues et l’histoire des conditions anthropiques environnantes.Enfin nous avons analysé les trajectoires évolutives suivies par un sous-ensemble de 12 sites, analysés à une échelle temporelle plus fine (5 observations par site sur 50 ans). Nous observons que les tronçons suivent des trajectoires différentes, en fonction des conditions hydrologiques observables à un moment donnée (ex. crues), de la position géographique du tronçon et du contexte anthropique. De plus, certaines conditions locales (topographie et humidité relative) semblent aussi nécessaire à la compréhension du patron biogéomorphologique des tronçons étudiés.A partir de ces résultats des éléments de perspectives pour une gestion intégrée des rivières en tresses du district rhodanien sont proposés. / The landscape of the braided rivers is characterised by a rich and diversified mosaic of habitats. The variability of the braided riverscape depends upon the combination of bio-physical factors, which interact at the different spatial and temporal scales.We did a comparative analysis at the regional scale of 53 braided reaches selected in the Rhône basin district. We used a remote sensing-based approach.At first, we analysed the aquatic habitat and the river channel pattern. We showed that the braiding and aquatic habitat patterns do not depend only on the flow regime but also on some local scale conditions, such us the sediment regime and the presence of the “near-floor” groundwater.Secondly, we moved on the overall riverscape corridor, and we applied a diachronic analysis comparing the present state of reaches with their state at the mid of the last century. The results show that not all studied braided reaches are narrower today compared to the 1950s and that some reaches are even wider. It is linked to a combination of several factors, such as the geographical position of a reach in its catchment and the short-term evolution ruled by recent large floods, as well as the history of the catchment linked to human settlements.Then we focused on the evolutionary trajectories followed by a sub-set of 12 braided reaches through time, and we increased the temporal resolution, for a total of five dates per reach. The results show that reaches follow different trajectories depending on: the hydrological conditions at a given period (e.g. floods), the geographical position of a reach at the regional scale and the anthropic context. Additionally, it seems that also local conditions (in terms of topography, sedimentation, and soil moisture), play an important role in determining the biogeomorphological pattern of the studied reaches.From these results, we discussed some perspectives for the braided river management in the Rhône basin district.
46

Rozpoznávání textu pomocí konvolučních sítí / Optical Character Recognition Using Convolutional Networks

Csóka, Pavel January 2016 (has links)
This thesis aims at creation of new datasets for text recognition machine learning tasks and experiments with convolutional neural networks on these datasets. It describes architecture of convolutional nets, difficulties of recognizing text from photographs and contemporary works using these networks. Next, creation of annotation, using Tesseract OCR, for dataset comprised from photos of document pages, taken by mobile phones, named Mobile Page Photos. From this dataset two additional are created by cropping characters out of its photos formatted as Street View House Numbers dataset. Dataset Mobile Nice Page Photos Characters contains readable characters and Mobile Page Photos Characters adds hardly readable and unreadable ones. Three models of convolutional nets are created and used for text recognition experiments on these datasets, which are also used for estimation of annotation error.
47

HISTORICKÝ VÝVOJ TECHNIKY PLAVECKÝCH ZPŮSOBŮ V SOUVISLOSTI SE ZMĚNAMI PRAVIDEL / THE HISTORICAL DEVELOPMENT OF THE TECHNIQUE OF SWIMMING STYLES IN CONNECTION WITH CHANGES IN THE RULES

Šatra, Michal January 2012 (has links)
Bibliographic identification ŠATRA, Michal: The historical development of the technique of swimming styles in connection with changes in the rules [Diploma thesis]. Charles University in Prague. Faculty of Physical Education and Sport. Department of Swimming. Supervisor: Mgr. Babeta Chrzanowská. Prague: FTVS UK, 2012. Abstract The aim of this thesis is to describe the development of the techniques of each individual swimming style on the basis of all available materials and to discover its connection with significant changes in the respective style's rules. The purpose of this thesis is to point out the most important changes in the rules of swimming within their historical context. A DVD which contains a practical demonstration of various swimming styles forms part of this work. Keywords Swimming, free style, breaststroke, backstroke, butterfly, swimming style, swimming technique, historical development, contemporary style, rules, historical photos, DVD.
48

Mesure de la section efficace différentielle de production du boson Z se désintégrant en paires électron-position, dans l'expérience ATLAS

Doan, Thi Kieu Oanh 28 November 2012 (has links) (PDF)
La première mesure du spectre en phi*_eta du boson Z à 7 TeV a été réalisée dans cette thèse. Cette variable permet de sonder la dynamique de production des Z de façon fine. L'échantillon complet des données enregistrées par ATLAS en 2011 a été utilisé ce qui correspond à 4.7/fb de luminosité intégrée. Les résultats de cette mesure sont publiés dans la Ref. [18] fondé sur la note interne Ref. [69]. La section efficace différentielle de Z->ee en fonction phi*_eta a été mesurée et comparée aux calculs perturbatifs à ordre fixé, avec/sans resommation pour la région des petits phi*_eta. Le code RESBOS fournit la meilleure description des données, cependant il est incapable de reproduire, à mieux de 4%, la forme détaillée de la section efficace mesurée. La section efficace différentielle a également été comparée aux prédictions de différents générateurs Monte Carlo interfacés avec un algorithme de parton shower. Les meilleures descriptions du spectre en phi*_eta mesuré sont données par les générateurs SHERPA et POWHEG+PYTHIA8. La mesure précise de la section efficace différentielle en phi*_eta fournit des informations précieuses pour l'ajustement des codes Monte Carlo. La précision expérimentale typique de cette mesure (~0.5%) est dix fois meilleure que la précision des calculs théoriques et elle est donc aussi précieuse pour contraindre la théorie. La mesure du spectre en ptZ a également été faite pour quantifier l'incertitude systématique de cette mesure en utilisant la grande statistique de l'échantillon de données. Cela permet de comparer deux mesures qui traitent de l'impulsion transverse du boson Z. Dans la plupart du domaine en phi*_eta l'incertitude systématique de la mesure de ptZ est deux fois plus grande que celle de la mesure de phi*_eta. Cette comparaison confirme l'intérêt de la variable phi*_eta. Les résultats présentés dans cette thèse ont beaucoup d'implications pour les études futures. Ajustant les générateurs Monte Carlo en utilisant les résultats de la mesure précise du spectre en phi*_eta minimisera l'incertitude sur leurs paramètres. Une mesure de la section efficace doublement différentielle en ptZ et phi*_eta est intéressante pour mieux comprendre la corrélation entre ces deux variables. La mesure précise du spectre en ptZ utilisant la variable phi*_eta peut être appliquée au spectre en ptW et on sait que des mesures plus fines du ptW sont importante pour une détermination précise de la masse du boson W. De plus, une compréhension précise du spectre en ptZ est importante pour comprendre les propriétés cinématiques de la production du boson de Higgs.
49

Βιοπαρακολούθηση (biomonitoring) των προστατευόμενων περιοχών του δικτύου Φύση 2000 με σύγχρονες μεθόδους τηλεπισκόπησης και τηλεμετρίας

Γεωργιάδης, Γεώργιος 20 April 2011 (has links)
Η επιστήμη της τηλεπισκόπησης και τα δορυφορικά δεδομένα, αποτελούν σήμερα ένα σύγχρονο επιστημονικό εργαλείο, αφού παρέχουν πλήθος σημαντικών πληροφοριών διευκολύνοντας σημαντικά την έρευνα σε πολλά επιστημονικά πεδία. Στην παρούσα εργασία παρουσιάζεται ένα ολοκληρωμένο σύστημα διαχρονικής παρακολούθησης της βλάστησης και των ανθρώπινων επιδράσεων στην προστατευόμενη περιοχή του Δάσους της Στροφιλιάς – λιμνοθάλασσας Καλογριάς και Κοτυχίου με σύγχρονες μεθόδους Τηλεπισκόπησης και Τηλεμετρίας. Ειδικότερα χρησιμοποιήθηκαν αεροφωτογραφίες και εικόνες από το δορυφόρο Landsat για απεικόνιση των φασματικών υπογραφών και ανίχνευση των μεταβολών της βλάστησης και του τοπίου. Επίσης χρησιμοποιήθηκαν δεδομένα από αισθητήρες μέτρησης υδρολογικών παραμέτρων για να δείξουμε χρονικές μεταβολές. Τέλος δημιουργήθηκε μία on-line βάση δεδομένων με σκοπό να διευκολύνει την εργασία πεδίου αλλά και το χρόνο της καταχώρησης της δειγματοληψίας. / In nowadays the science of remote sensing and satellite data, represent a modern scientific tool, due to the fact that can provide plenty of important information making easier the research in a lot of scientific fields. In the present work we demonstrate a complete biomonitoring system of diachronic follow-up of vegetation and human effects in the protected region of Strofylia's Forest and the lagoons of Kalogrias and Kotyhi, with modern remote sensing methods and Telemetry. More specifically we used aerial photos and satellite images from the Landsat satellite in order to illustrate spectrum signatures and detection of changes. Also we analyzed data from sensors that measured environmental parameters in order to show time changes. Finally we created an online database so as to store all this important information that we produced but also to minimize the total time of registration and sampling.
50

Explorando a localização e orientação de fotografias pessoais para descoberta de pontos de interesse baseada em agrupamento.

LACERDA, Yuri Almeida. 16 May 2018 (has links)
Submitted by Kilvya Braga (kilvyabraga@hotmail.com) on 2018-05-16T14:43:46Z No. of bitstreams: 1 YURI ALMEIDA LACERDA - TESE (PPGCC) 2017.pdf: 28353805 bytes, checksum: 0d8453f39b64823b7308c81cc7731524 (MD5) / Made available in DSpace on 2018-05-16T14:43:46Z (GMT). No. of bitstreams: 1 YURI ALMEIDA LACERDA - TESE (PPGCC) 2017.pdf: 28353805 bytes, checksum: 0d8453f39b64823b7308c81cc7731524 (MD5) Previous issue date: 2017 / Capes / A descoberta de conhecimento a partir de grandes repositórios online de fotografias tem sido uma área de pesquisa bastante ativa nos últimos anos. Isso se deve principalmente a três fatores: incorporação de câmeras digitais e sensores de geolocalização aos dispositivos móveis; avanços na conectividade com a Internet; e evolução das redes sociais. As fotografias armazenadas nesses repositórios possuem metadados contextuais que podem ser utilizados em aplicações de descoberta de conhecimento, tais como: detecção de pontos de interesse (POIs); geração de roteiros de viagens; e organização automática de fotografias. A maioria das abordagens para detecção de POIs parte do princípio que as áreas geográficas onde uma grande quantidade de pessoas capturou fotografias indica a existência de um ponto de interesse. Porém, em muitos casos, os POIs estão localizados a uma certa distância desse local na orientação em que a câmera estava direcionada, e não no ponto exato da captura da fotografia. A maioria das técnicas propostas na literatura não consideram o uso da orientação no processo de detecção de pontos de interesses. Dessa forma, este trabalho propõe novos algoritmos e técnicas para detecção de pontos de interesse em cidades turísticas a partir de coleções de fotografias orientadas e georreferenciadas explorando de diversas formas a orientação geográfica. Esta pesquisa comprovou a importância do uso da orientação nos novos algoritmos voltados para detecção de pontos de interesses. Os experimentos, utilizando uma base de dados real de grandes cidades, demonstraram que os algoritmos considerando a orientação conseguem, em alguns cenários, superar os que não a consideram. Também foram propostas novas métricas de avaliação e uma ferramenta para auxiliar as atividades de descoberta de conhecimento baseada em grandes massas de fotografias. / The knowledge discovery from huge photo repositories has been a very active area of research in the last years. This is due to three facts: the incorporation of digital cameras and geolocation sensors in mobile devices; the advances in Internet connectivity; and the evolution of social networks. The photos stored on those repositories have contextual metadata. Those metadata could be used for many applications of knowledge discovering, such as: Point of Interest (POI) detection; generating of tourist guides; and automatic photo organization. Most approaches for POI detection assume that geographic areas with high density of photos indicate the existence of a point of interest in that area. However, in many cases, the POIs are located in a certain distance of that position in direction where camera was aiming, and not in the exact point of photo shooting. Most of related work do not consider the use of orientation in the process of POI detection. In this way, we propose a set of algorithms and techniques for POI discovery in touristic cities using geotagged and oriented photos collection exploring the geographic orientation in different ways. This research has proven the importance of the usage of orientation in the new algorithms for POI detection. In the experiments with collections related to big cities, the algorithms considering orientation, in several scenarios, have beaten those that do not consider. Also, new metrics of evaluation have been proposed and a new framework to assist all the tasks for knowledge discovery based on huge photo collections.

Page generated in 0.0461 seconds