Šio darbo tikslas – sukurti prototipą, kurio pagalba būtų galima atlikti efektyvų viename kompiuteryje esančių skatmeninių vaizdų paieškos rezultatų pateikimą subjektui. Prototipo veikimas yra paremtas skaitmeninių vaizdų įverčių skaičiavimo algoritmu. Pagrindinis akcentas – tai skaitmeninių vaizdų paieškos rezultatų vertinimas bei prioretizavimas. Prioretizacija vykdoma apskaičiuojant kiekvienam skaitmeniniui vaizdui įvertį, kuris nurodo pastarojo svarbą rezultatų aibės kontekste. Įverčio paieška vykdoma naudojant panašumo grafą bei taikant jame tikrinio vektoriaus centriškumo algoritmą. Darbo metu aptarti algoritmo privalumai bei trūkumai. Pateikiamos prielaidos, kurios leistų optimizuoti pastarojo veikimą. Optimizacija yra paremta daugiamačių duomenų indeksavimo spartos gerinimu. Taip pat siūlomas techninis sprendimas leidžiantis efektyviai naudoti sukurtąjį prototipą realioje operacinėje sistemoje. Darbe siūloma naudoti KD medžius. Tai alternatyva Henry Rowly, Shumeet Baluja ir Yushi Jing pasiūlytiems kamuoliniams medžiams. Darbo rezultatas – prototipas, kurio pagalba galima atlikti efyktyvų skaitmeninių vaizdų paieškos rezultatų pateikimą subjektui. Pasinaudojant pastarąja prgrama atlikta praktinė analizė. Tyrimo metu nustatyta, kad KD medžių panaudojimas ne tik supaprastina algoritmą, bet ir leidžia įvykdyti viename kompiuteryje esančių skaitmeninių vaizdų paiešką greičiau nei naudojant kamuolinius medžius, o sukurtojo prototipo panaudojimas leidžia efektyviai... [toliau žr. visą tekstą] / In this paper the problem of image search is addressed. The primary goal of the paper is to present a prototype that can be used to identify and effectively return search results to the end user (i.e., return the most relative images to some given search criteria). All the search is performed to find images stored in one personal computer. To accomplish this task an image rank algorithm is utilized. Algorithm's advantages and disadvantages are pointed out. As a result an optimization issue is discussed and a way to increase the performance of multidimensional data indexing is presented. The optimization is based on KD trees. The assumption is that these trees might work faster than ball trees. Those were proposed by authors of the original image rank algorithm. The result is a prototype that is used for searching and presenting images based on ranks that indicate relative importance of each of the image to some given search criteria. The results are prioritized. The prioritization might be thought of as sorting images by their similarities. By using the application a practical analysis was performed and the results were described in details. It was found that KD trees can be effectively used to index multidimensional data while calculating image ranks. It not only simplifies the algorithm but also provides better performance results than ball trees.

The use of advanced technologies in medicine allows to diagnose diseases in their initial stage. Its automatical parameterization is possible only in clear, qualified pictures. The paper investigates the initial processing of the picture of eye bottom which is necessary for the parameterization of sight nerve disc. This is the change of a colourful picture into a grey one and covering the picture with a net what allows to find a sight nerve disc. The paper describes the problems which disturb to precisely localize a sight nerve disc. This would be a net of blood-vessels, which is treated as noise in pictures. There is also a shortly discussed possible method for the further picture treatment – parameterization of a sight nerve disc. The research investigates and looks for optimal netting parameters that could more precisely and faster to find a sight nerve disc. A realised programme is presented which performs the initial picture treatment allowing to choose necessary netting parameters: the size of a cell, pace and form. In addition, the instruction of programme using is presented. The results received during the research are illustrated with the experiment description, diagrams and remarks.

Uncompressed multimedia data requires considerable storage capacity and transmission bandwidth. Despite rapid progress in mass-storage density, processor speeds, and digital communication system performance, demand for data storage capacity and data-transmission bandwidth continues to outstrip the capabilities of available technologies. The recent growth of data intensive multimedia-based web applications even more sustained the need for more efficient ways to encode such data. There are two types of image compression schemes – lossless and lossy algorithms. In lossless compression schemes, the reconstructed image, after compression, is numerically identical to the original image. However lossless compression can only achieve a modest amount of compression. An image reconstructed following lossy compression contains degradation relative to the original. Often this is because the compression scheme completely discards redundant information. However, lossy schemes are capable of achieving much higher compression. The aim of this research is to create an efficient lossy image compression algorithm, using heuristic data clusterization methods; perform experiments of the new algorithm, measure its performance, analyze advantages and disadvantages of the proposed method, propose possible improvements and compare it with other popular algorithms. In this paper is presented new algorithm for image compression, which uses data base of popular image fragments. Proposed algorithm is... [to full text]

One of the images compression methods – fractal image compression is analyzed in the work. After work carried out, it is possible to state, that selecting parameters of method of fractal compression depends on user’s demands.

The aim of the work. The aims of this work are: a) to develop new or modify existing face detection, analysis and recognition methods, in order to increase their speed and accuracy; b) investigate the problems of face features and face contour detection.

Šiame darbe pateikiamas apibendrintas automatinio vaizdų jungimo į panoramas algoritmas, detaliai aptariami algoritmo žingsniai ir galimas našumo didinimas lygiagretaus programavimo priemonėmis. Tyrime aprašyti vaizdų jungimo bandymai panaudojant kelias populiarų vaizdų bruožų alternatyvas. Taip pat pristatoma praktinė vaizdų jungimą panaudojanti aplikacija. / This study presents a generalization of an automated image stitching algorithm, describes it's steps in great detail and discusses possible performance improvements by use of parallel execution. Results of stitching experiments using a couple of modern and popular image feature alternatives are shown. A practical application using an implementation of automated image stitching is presented.

Vaizdų segmentavimas yra naudojamas daugelyje mokslo sričių, tokių kaip geografija, geologija ar biologija, tam, kad galėtume atskirti bei išanalizuoti objektus apdoruojamuose paveikslėliuose. Rinkdamiesi vaizdų segmetavimo algoritmą dažniausiai atkreipiame dėmesį į jo efektyvumą, sudėtingumą, vykdymo laiką bei pritaikymą skirtingiems vaizdų tipams. Kadangi dažniausiai segmentuojame sudėtingos struktūros bei blogesnės kokybės vaizdus, labai svarbu, kad segmentavimo algoritmas tiktų visų klasių vaizdams. Mumford Shah funkcionalas yra laikomas vienu populiariausių ir efektyviausių vaizdų segmentavimo įrankių. Algoritmas pasižymi lengva realizacija programoje bei greitu įvykdymu. Didžiausiu šio algoritmo privalumu yra galimybė pritaikyti jį skirtingo tipo vaizdams. Keisdami į Mumford Shah funkcionalo formulę įeinančių parametrų reikšmes galime įtakoti segmentavimo rezultatus. Darbe yra išsamiai išnagrinėjami visi įmanomi funkcionalo formulės atvejai su skirtingomis parametrų reikšmėmis ir bandoma išanalizuoti, kokios parametrų reikšmės geriausiai tinka atskiroms vaizdų klasėms. Visi gauti rezultatai ir tezės yra pagrindžiami statistiniu tyrimu. / Image segmentation is used in many fields of science such as geography, geology and biology, so that we can distinguish and analyze objects in pictures. When selecting segmentation algorithm we usually draw attention to its performance, complexity and adaptation to different types of images. As we usually are processing complex structure and lower-quality images, it is very important that the segmentation algorithm would be suitable for all classes of images. Mumford Shah functional is one of the most popular and effective tools for image segmentation. The algorithm is characterized by easy and fast realization. The biggest advantage of this algorithm is the ability to adapt to different types of images. By changing the Mumford Shah functional formula‘s incoming parameter‘s values, we can affect the segmentation results. In this work we are selecting the most suitable parameter‘s values for different types of images in order to obtain the best segmentation results. Obtained results are based on statistical analysis.

The value of information often depends on how easy it can be found, retrieved, accessed, filtered and managed. This challenging situation demands a timely solution to the problem. MPEG-7 is the answer to this need. MPEG-7, formally named “Multimedia Content Description Interface”, is a standard for describing the multimedia content data that supports some degree of interpretation of the information’s meaning, which can be passed onto, or accessed by, a device or a computer code. The main objective of this project is to create low complicated, small size, more concrete system, which main features are finding, retrieval, access, filtration and management of broad range of pictures.

This Master degree paper analyses hyperbolic image filtering in spaces of different dimensionality. It investigates the problem of optimal filtering space selection. Several popular image compression methods (both lossless and lossy) are reviewed. This paper analyses the problems of image smoothness parameter discovering, image dimensionality changing, hyperbolic image filtering and filtering efficiency evaluation and provides the solution methods of the problems. Schemes for the experimental examination of theoretical propositions and hypotheses are prepared. This paper comprehensively describes experiments with one-, two- and threedimensional images and the results of the experiments. Conclusions about the efficiency of hyperbolic image filtering in other than "native" image space are based on the results of the experiments. The criterion for the selection of optimal image filtering space is evaluated. Guidelines for further research are also discussed. The presentation Specific Features of Hyperbolic Image Filtering, which was based on this Master degree paper, was made at the conference Mathematics and Mathematical Modeling (KTU – 2004). This text is available in appendixes.

Baigiamajame magistro darbe nagrinėjami struktūrų modeliavimo, diskretizavimo ir optimizavimo uždaviniai, jų sprendimo būdai ir algoritmai. Pasiūlyta originali strypinių struktūrų optimizavimo ir diskretizavimo technika, kurioje naudojami vaizdų algebros metodai ir baigtinių elementų metodo (toliau – BEM) programiniai paketai. Atlikta BEM programinių paketų apžvalga, parinkti tinkamiausi paketai darbo tikslams realizuoti ir rezultatams verifikuoti. Išanalizuoti skaitmeninių vaizdų skeletų išgavimo metodai. Pasiūlytas originalus vaizdo skeleto apdorojimo algoritmas skerspjūvio charakteristikoms nustatyti. Sudarytas ir programiškai realizuotas strypinių struktūrų optimizavimo-diskretizavimo algoritmas. Realizuota programinė sistema susideda iš vaizdų apdorojimo ir duomenų paruošimo dalies (MATLAB kalba) bei BEM skaičiavimų dalies (ANSYS vidine programavimo kalba APDL). Skaičiavimo rezultatai atvaizduojami ir verifikuojami STAAD.Pro paketu. Diskretizavimo metu strypinės struktūros mazgų vieta randama vaizdo skeleto segmentų sankirtos taškuose, o segmentų skerspjūvių plotai randami atkarpų, jungiančių šiuos mazgus, vidurio taškuose. Kai struktūros mazgų padėtis fiksuota arba mazgai yra per arti vienas kito atliekamas mazgų padėties koregavimas. Darbe atlikti testiniai skaičiavimai, rezultatų analizė ir verifikavimas, suformuluotos išvados. / In the master thesis the problems of structure modeling, discretization-optimization and their solution methods and algorithms are analyzed. The original technique for optimization and discretization of beam structures has been suggested; The packages of image algebra methods and of the finite element methods were employed for that. Several packages of finite element method have been reviewed and the most suitable packages for the current problems were identified. The methods for obtaining skeletons of digital images were explored. The algorithms for optimization and discretization of beam structures has been suggested and coded. The program created consents of the part for image processing and input data preparing, and the part for image the finite element via method. The results obtained are represented and verified by STAAD.Pro package. During the discretization, the positions of structure nodes are obtained in the intersection of skeleton segments. The segments' cross-section areas are obtained in the middle-points between two adjacent nodes. The positions of nodes may be corrected if the nodes close to each other. The test-calculation, analysis of results and verification are presented and conclusions are drawn.