Now showing items 1-20 of 579
Next Page| Abstract: | Images are commonly used on a daily basis for research, information and entertainment. The introduction of digital cameras and especially the incorporation of cameras in mobile phones makes people able to snap photos almost everywhere at any time since their mobile phone is almost always brought with them. The fast evolution in hardware enables users to store large image collection without high costs. Making use of these image collections requires efficient image retrieval techniques. Traditional image retrieval techniques like text-based image retrieval and content-based image retrieval techniques have shortcomings. New techniques or combination of existing techniques must be established to provide users with adequate image retrieval functionality. This thesis describes two systems enabling users to retrieve information such as images, textual information, WAP-links or videos using SMS or MMS. One of the services, M2S is meant for tourists to retrieve information about attractions in Lofoten. M2S uses content-based image retrieval to retrieve the information requested. This service is designed and implemented in cooperation with Telenor R&I. The other system, CAIR is meant for users who want to retrieve images from an image collection using SMS. CAIR uses a context-based image retrieval to retrieve images. This system is designed, but not yet implemented. |
| URI: | http://hdl.handle.net/10037/1141 |
| Abstract: | In recent years, an increasing amount of personal images, video, sound and text data are captured and stored in a digital format. Increased storage capacity at lower cost entice us to attempt to store everything, but without effective information retrieval techniques, the usefulness of the data becomes limited. Some people have taken personal data capture to extremes and have begun to capture digitally all aspects of their life, which creates enormous archives of multimedia data. This is not a new idea: in 1945 Vannevar Bush wrote his visionary article “As We May Think” where he described Memex, the first Human Digital Memory (HDM). Today we have projects like Microsoft MyLifeBits building on the Memex vision, however there has been little focus on organizing this kind of data effectively. By applying data reduction, we show the benefits of removing redundancy from HDMs, and illustrate how the same data reduction framework can be used to effectively support information access from HDMs. |
| URI: | http://hdl.handle.net/10037/1825 |
| Abstract: | Existing networked filesystems are usually either client/server - allowing storage only on one node - hard to use, or both. The advanced ones also like to use their own on-disk format, complicating migration both ways. Skynet attempts to remedy this. It is a distributed filesystem with master/slave redundancy that is easy to use, relatively safe for your data and can be easily converted to/from a non-distributed filesystem. It uses an existing filesystem for file storage, takes care of its own maintenance as far as possible, and supports approximate POSIX semantics, with POSIX, eventual and session coherency modes. It favors speed over correctness, where this would rarely be noticed. Part of the project includes a cryptographically secure message-passing middleware for Haskell called Hermes, with distribution transparency and a gossip system. High-level SHA2 and AES bindings are also included. The Hermes portion of the project is complete and usable, the Skynet portion is not. Skynet and Hermes are designed for small-to-medium networks, achieving optimal performance at this size and suffering significant degradation in large networks. They are meant to be used for non-administrated applications in home networks. Skynet offers single-bit security. The network is encrypted and authenticated, but there is no further security inside the network. |
| URI: | http://hdl.handle.net/10037/2543 |
| Abstract: | Clustering is an unsupervised pattern recognition technique for finding natural groups in data, whether it is a grouping of web pages found by a search engine or segmenting satellite images into different types of ground cover. There exists a variety of different ways to perform clustering ranging from heuristics rules designed for a specific dataset to general procedures which can be applied to all datasets with varying degrees of success. The k-means algorithm is a well known example of the latter approach that can be expected to give good results when the data is easily separable. Another example of general clustering procedures are spectral clustering methods which involve a non-linear data transformation that allows them to handle complex cluster structures. They are considered to be the state of the art in the clustering literature, but are computationally demanding and unable to handle large datasets. To overcome the size limitations, this thesis uses a two-stage approach. The first stage can be seen as a preprocessing to reduce the size of the input for the spectral clustering in the second stage. The preprocessing is accomplished by using a simple clustering method on the dataset. These clusters represents a partitioning and is a more natural way of representing the dataset than randomly selecting a subset of samples. One can then adjust the number of partitions so spectral clustering methods can be used. The procedure is called Mean Shift Spectral Clustering (MSSC) as the mean shift clustering algorithm is used in the first stage. Each partition found by mean shift consists of data points that are close to the same mode in the estimated probability density function. Hence the partitions will be representative of the geometric structure of the dataset. This thesis realizes for the first time the idea of spectral clustering based on the partitions found by mean shift. The Kernel Entropy Component Analysis (KECA) spectral clustering method, a recent development in the field of spectral clustering, is used for this purpose and compared with the better known Kernel Principal Component Analysis (KPCA) method. A comprehensive collection of MATLAB functions has been developed to allow the testing of this procedure which is able to handle large and high dimensional datasets. It is able to cluster images directly with as many feature vectors as there are pixels. The experiments show how the clustering accuracy varies as a function of the primary parameters. This gives a good overall characterization of the method and what can be expected when used for unfamiliar datasets. The MSSC procedure is shown to provide good clustering results when following some basic principles for selecting parameters. |
| URI: | http://hdl.handle.net/10037/4610 |
| Abstract: | My master thesis concentrates on the distribution of gas accumulation to infer vertical fluid migration in sedimentary formations of Eocene-Oligocene (Brygge Formation) and Pliocene-Pleistocene (Naust Formation) in the mid-Norwegian margin at the Helland Hansen Arch, north of the Storegga slide. The seismic interpretation is based on “Helland Hansen 3D cube SH9602”. The N-S oriented Helland Hansen Arch (HHA) has a hydrocarbon potential where both tectonic and thermal subsidence has played a role in forming hydrocarbon traps. The oozes and debris flow deposit seal off the vertical migrating fluids allowing them to accumulate along the HHA. At the crest of HHA seismic bright spots are delineated by using acoustic attributes. The area beneath the bright spots shows the depletion of higher frequencies due to absorption of seismic energy. Beneath the bright spot vertical fluid migration features appear as disturbed and distorted wipe-out seismic zone and velocity sags (push down). The vertical fluid migration features occur along major faults and are identified by the low coherency values on variance attribute maps. Gas accumulation zones are extensively disturbed by faulting along the eastern flanks of HHA where polygonal faults provide routes for vertical gas migration. The polygonal faults were reactivated by sediment loading from submarine slides. Pipe structures provide further pathways for vertical gas migration from reactivated polygonal faults. The presence of gas in the glacigenic sedimentary formation of Naust A and U unit is identified by the distinct reduction of P-wave velocities. |
| URI: | http://hdl.handle.net/10037/4320 |
| Description: | The papers of this thesis are not available in Munin: 1. Said H. Ahmed and Tor Flå: 'Estimation of Evolutionary Average Hydrophobicity Profile from a Family of Protein Sequences', Lecture Notes in Computer Science (2007) Volume 4774/2007, 158-165. Available at http://dx.doi.org/10.1007/978-3-540-75286-8_16 2. Said H. Ahmed and Tor Flå: 'Evolutionary Parameters in Sequence Families : Cold adaptation of enzymes', Lecture Notes in Computer Science (2009) Volume 5780/2009, 1-12. Available at http://dx.doi.org/10.1007/978-3-642-04031-3_1 3. Tor Flå and Said H. Ahmed: 'Evolution of cold-adapted protein sequences', in Sequence and Genome analytics : methods and application II (ed. Gabriel Fung), Concept Press Ltd (2010), ISBN 9780-9807330-6-8 4. Said H. Ahmed and Tor Flå: 'Position dependent mean hydrophobicity and structural profiles' (unpublished paper) |
| URI: | http://hdl.handle.net/10037/3533 |
| Abstract: | This study presents the weighted least squares method based on Interferometric Synthetic Aperture Radar (InSAR) images to retrieve spatial-temporal evolution of land subsidence in Mashhad Valley, northeast Iran. Using the analysis of a few interferograms covering the 2003-2005 period, Motagh et al (GJI 2006) presented a preliminary analysis of the subsidence in this area. Here we extend this study and use additional SAR data to retrieve time-dependent deformation in Mashhad Valley We utilize 17 SAR images acquired by the ENVISAT satellite in a descending orbit during Jun. 2004-Nov. 2007, make 53 differential interferograms spanning different long- and short-term intervals, and do a time series analysis to extract deformation signals out of differential interferograms. Time series analysis suggests that the subsidence occurs within a northwest-southeast elongated elliptical shaped bowl along the axis of Mashhad valley. The maximum accumulated subsidence during 1260 days reaches about 86 cm, located northeast of Mashhad City. The comparison between InSAR time series results with continuous GPS station in the city of Tous, northeast of Mashhad, yields comparable results at the level of 1 cm. |
| URI: | http://hdl.handle.net/10037/3066 |
| Abstract: | Dette er en eksperimentell oppgave der et bredbåndet ultralydsystem er brukt til å undersøke hvordan dempingen av ultralydbølger i elastiske materialer, avhenger av frekvensen. For å gjøre denne undersøkelsen ble det utviklet en holder i pleksiglass med en ultralydtransduser på hver side. I denne holderen kan prøver av ulike materialer lett festes og undersøkes ved at holderen plasseres i et vannbad. Det eksperimentelle oppsettet muliggjør målinger av både reflektert og transmittert lyd. Som målesystem ble det brukt et såkalt ”pulser-receiver” system for ultralyd, som leveres som et PCI-kort for plassering i en standard PC. |
| URI: | http://hdl.handle.net/10037/2088 |
| Abstract: | In 2008 Scandinavian Highlands discovered a gold mineralization within the Mauken basement window. The Mauken basement window is a thought to be an eastward continuation of the West Troms Basement Complex which is a series of rocks with an age span from Neoarchaean to Paleoproterozic. The mineralization found in a steeply dipping NW-SE striking shear zone and is hosted in a meta-sedimentary unit. The rock which hosts the mineralization is metamorphosed under greenschist facies conditions. The protolith for the mineralization has not been determined and could be of magmatic or of sedimentary origin, which the rocks surrounding the mineralization are. Surrounding the mineralization is a halo of sericitization, but the rock is also influenced by chloritization and local carbonatization. The mineral assemblage of the mineralization is dominated by quartz, micas, and carbonates. The gold mineralization is closely linked to the formation of arsenopyrite, since most of the gold is found as inclusions in arsenopyrite grains. The deposit is likely to have been formed in relation to the Svecofennian orogeny and could be classified as an orogenic gold deposit. |
| URI: | http://hdl.handle.net/10037/5209 |
| Abstract: | Absolute binding free energies for the third domain of the turkey ovomucoid inhibitor in complex with Streptomyces griseus proteinase B and porcine pancreatic elastase has been calculated using the linear interaction energy method. |
| URI: | http://hdl.handle.net/10037/3820 |
| Abstract: | OOPP is a component based middleware platform with support for complex distributed applications. The main goal of OOPP is to create an expressive programming model for distributed applications where by default details are hidden for the programmer. When necessary, reflection is used to expose and sometimes modify these details. All interaction with an OOPP component are specified by its component model and a at set of well-defined interfaces. The component model specifies how a component interacts with the runtime. The set of interfaces specifies how a component interacts with other components. An OOPP application is created by combining a set of OOPP components. OOPP Studio can be used to create and manage such an application. This paper gives an overview of the OOPP programming model and discusses in more details OOPP Studio. OOPP Studio is a powerful graphical tool used to build, deploy, and manage a distributed OOPP application. |
| URI: | http://hdl.handle.net/10037/390 |
| Abstract: | Ultrabredbåndet (ultrawideband - UWB) radar brukt til deteksjon av bryst-kreft viser oppløftende resultater på grunn av dens gode romlige oppløsning og dype penetreringsdybde. Bakgrunnen for å bruke UWB-radar til dette formålet er kontrasten mellom de dielektriske egenskapene til friskt brystvev og en kreftsvulst. Noe som vil føre til relativt mye reflektert energi fra kreftsvulsten. En av utfordringene med UWB-radar brukt til deteksjon av brystkreft er å få transmittert mest mulig av signalenergien inn i brystet og fram til en eventuell kreftsvulst. For å øke dette signal-til-støyforholdet er det foreslått å bruke aktiv stråleforming for å fokusere signalenergi i bestemte posisjoner. Dette er tidligere utprøvd i numeriske simuleringer med oppløftende resultater. I denne oppgaven gjøres det eksperimentelle forsøk for å undersøke om aktiv stråleforming, over så korte avstander, vil øke signal-til-støyforholdet også i praksis. Det etableres et eksperimentelt oppsett til bruk av UWB-radar. Det brukes et antenne-array med fire antenner, og som brystvevfantom brukes det soyaolje. For å generere signal brukes det en nettverksanalysator som sender ut en syntetisk UWB-puls. Denne nettverksanalysatoren brukes også til å måle det elektriske feltet ved hjelp av en probe. For å sikre et mest mulig korrekt bilde av det elektriske feltet, karakteriserer vi proben for å kunne kompensere for dens påvirkning på måleresultatene. Soyaoljen karakteriseres også, dette for å kunne utføre stråleformingen korrekt. For å studere effekten av aktiv stråleforming måles strålingsdiagrammet til antenne-arrayet for fokusering i ulike posisjoner. Det utføres også numeriske simuleringer av modeller som er mest mulig lik de eksperimentelle oppsettene, det gjør at vi kan sammenligne resultatene fra de to tilfellene direkte. Resultatene av de eksperimentelle forsøkene viste at aktiv stråleforming øker energien og dermed signal-til-støyforholdet, i fokuspunktene, betraktelig. Nøyaktigheten til oppnådd fokuspunkt i forhold til ønsket var også god. Det samme ble oppnådd for de numeriske simuleringene, og overenstemmelsen mellom strålingsdiagrammene oppnådd ved eksperimentelle målinger og numeriske simuleringer var god. |
| URI: | http://hdl.handle.net/10037/2090 |
| Abstract: | Oppgaven omhandler målemetoder og problemstillinger i forhold til temperaturmåling på optiske bølgeledere |
| URI: | http://hdl.handle.net/10037/1203 |
| Description: | Innlegg på seminaret "Science - who cares?" ved NT-fak, Universitetet i Tromsø. |
| URI: | http://hdl.handle.net/10037/2695 |
| Abstract: | This work discusses the theory of the magnetizability. The magnetizability is a molecular property defined as the second-order perturbation of the energy in the presence of an external magnetic field. When the magnetic field is static, perturbation and multipole theories lead to an origin-independent expression for the magnetizability, whereas for a frequency-dependent magnetic field its expression turns out to be origin dependent. Since the magnetizability in the static case is an observable quantity, the expression obtained in the dynamic case is clearly unphysical. My work discusses several approaches in order to solve this issue. The magnetizability has been analyzed in the microscopic case, within both the non-relativistic and the relativistic framework. The expression for the magnetizability proposed by Raab and de Lange has been considered, pointing out the weak points of their non-relativistic theory, and trying to recover an alternative expression in the relativistic framework. The macroscopic case has also been discussed in detail, and an origin-independent expression for the inverse permeability has been derived within a fully relativistic approach. The problem of finding an origin-independent expression for the magnetizability in the microscopic case remains unresolved, however the results obtained in the macroscopic case and the analysis done in this thesis are good starting points for further studies, and suggests that future work should start from the relativistic theory of the property. |
| URI: | http://hdl.handle.net/10037/2961 |
| Abstract: | This thesis presents methods for statistical analysis of the probability distributions used to model multilook polarimetric radar images. The methods are based on a matrix-variate version of Mellin's integral transform. The proposed theoretical framework is referred to as Mellin kind statistics. It is an extension of a theory recently developed for single polarisation amplitude and intensity data to the complex matrix-variate case describing multilook polarimetric images. This generalisation is made possible by the rediscovery of a generalised Mellin transform, which is defined for functions of positive definite Hermitian matrices. The domain makes it suited for application to the distributions used to model the polarimetric covariance and coherency matrix. The analysis tools include the matrix-variate Mellin kind characteristic function, which is defined with the Mellin transform in place of the conventional Fourier transform. Matrix log-moments and matrix log-cumulants are retrieved from this function. The matrix log-cumulants are used in a moment based approach to parameter estimation of the distribution parameters. The estimators make efficient use of all the statistical information in the polarimetric covariance matrix, and are superior to all known alternatives. The matrix log-cumulants are also used to construct the first known goodness-of-fit test for matrix distributions based on the multilook polarimetric product model. The algorithms are interpreted by means of a highly informative graphical visualisation tool displaying a space spanned by certain matrix log-cumulants. It is demonstrated that the matrix-variate Mellin transform is the natural tool for analysing multilook polarimetric radar images. This conclusion is based on the simple and elegant mathematical expressions obtained, the superb statistical properties of developed estimators, as well as the intuitive interpretations offered by the Mellin kind statistics. |
| URI: | http://hdl.handle.net/10037/2471 |
| Abstract: | I klassisk digital kommunikasjon overføres en bitstrøm gjennom en kanal ved å modulere parametrene til en deterministisk bærebølge. Noen kjente eksempler er amplitudemodulasjon (AM), frekvensmodulasjon (FM) og fasemodulasjon (PM). Mottakeren estimerer parametrene til det informasjonsbærende signalet og bruker en deteksjonsregel til å klassifisere den mottatte bølgeformen som en av flere mulige parametriske signaler. Konvensjonelle metoder gir ingen beskyttelse mot at ikke-autoriserte mottakere avlytter og dekoder signalet. Blant ferske metoder som tilbyr en viss beskyttelse mot avlytting finner vi såkalte spredt spektrum-teknikker og kaotiske digitale kodere. Slike teknikker krever nøyakig synkronisering mellom sender og mottaker. Selv små synkroniseringsfeil kan føre til store bitfeilrater hos mottakeren. Denne oppgaven har undersøkt et nytt konsept innen digital kommunikasjon, som er basert på å bruke realisasjoner av stokastiske prosesser som informasjonsbærende signaler. Teknikken har innebygd sikkerhet mot avlytting. Samtidig er det mulig å konstruere dekodere som er enklere enn de som brukes i spredt spektrum og kaotisk kommunikasjon. Målet med prosjektet er å undersøke noen fundamentale egenskaper ved den nye teknikken: 1) Hvordan kan vi finne et statistisk mål på avstand mellom de informasjonsbærende stokastiske prosessene? 2) Hvilken detektor skal vi bruke for å dekode informasjonssekvensen, og hvor god ytelse gir denne? 3) Hvordan skal vi velge de stokastiske prosessene? |
| URI: | http://hdl.handle.net/10037/235 |
| Abstract: | A distributed file repository is described. It supports interaction between different machines used by a single user, as well as between users that share data. Files can be replicated and consistency will be maintained, or files can be shipped (copied) to a remote site. As with more traditional systems, the servers are trusted not to leak information. However, the rôle servers play is not as much the hub in the system. In particular, users are in charge of delegating acccess to files. For flexibility, delegations might take place outside of the realm of the system proper; by any means available to the users. Users can delegate access rights to local and remote users, including remote users in other domains. ACLs are used to maintain local access control; capabilities are used to access remote files. These capabilities are valid within epochs, but are immediately revoked when being used, thus being valid at-most-once. In essence, we have realized a flexible infrastructure where users can implement their own security policy. |
| URI: | http://hdl.handle.net/10037/360 |
Now showing items 1-20 of 579
Next Page
Munin is powered by DSpace 1.8.2
The University Library of Tromsø, N-9037 Tromsø
Tel: +47 77 64 40 00, E-mail: munin@ub.uit.no