|
|
Image Super-Resolution Using Deep Learning
Bublavý, Martin ; Juránková, Markéta (oponent) ; Španěl, Michal (vedoucí práce)
The ability to identify and treat a variety of medical diseases is made possible by medical imaging, which is an essential component of contemporary healthcare. Yet, elements like noise and low resolution can have a negative impact on the quality of medical photographs. In this thesis, how to enhance the resolution and quality of medical images was investigated using MedSRGAN, a deep learning model built on generative adversarial networks (GANs). MedSRGAN was implemented and then applied to computed tomography (CT), one of the most utilized medical imaging methods.
|
|
|
Evaluation of thin discontinuities in planar conducting materials using the diffraction of electromagnetic field
Savin, A. ; Nový, F. ; Fintová, Stanislava ; Steigmann, R.
The current stage of nondestructive evaluation techniques imposes the development of new electromagnetic (EM) methods that are based on high spatial resolution and increased sensitivity. In order to achieve high performance, the work frequencies must be either radifrequencies or microwaves. At these frequencies, at the dielectric/conductor interface, plasmon polaritons can appear, propagating between conductive regions as evanescent waves. In order to use the evanescent wave that can appear even if the slits width is much smaller that the wavwelength of incident EM wave, a sensor with metamaterial (MM) is used. The study of the EM field diffraction against the edge of long thin discontinuity placed under the inspected surface of a conductive plate has been performed using the geometrical optics principles. This type of sensor having the reception coils shielded by a conductive screen with a circular aperture placed in the front of reception coil of emission reception sensor has been developed and 'transported' information for obtaining of magnified image of the conductive structures inspected. This work presents a sensor, using MM conical Swiss roll type that allows the propagation of evanescent waves and the electromagnetic images are magnified. The test method can be successfully applied in a variety of applications of maxim importance such as defect/damage detection in materials used in automotive and aviation technologies. Applying this testing method, spatial resolution can be improved.
|
|
|
The Impact of Image Resolution on the Precision of Content-based Retrieval
Navrátil, Lukáš ; Lokoč, Jakub (vedoucí práce) ; Skopal, Tomáš (oponent)
Tato práce se zaměřuje na porovnání metod pro podobnostní vyhledávání obrázků. Jsou představené používané techniky a jsou ukázány testovací sady, které slouží k meření přesnosti vyledávacích systémů založených na podobnostním vyhledávání obrázků. Na modelech, implementovaných na základě představených technik, jsou pak prováděna různá měření, která zkoumají jejich výsledky v závislosti na vstupních datech, použitých komponentách a nastavení parametrů, přičemž speciální pozornost je věnována chování modelů pro různá rozlišení obrázků. Tyto výsledky jsou dále analyzovány a jednotlivé modely jsou vzájemně porovnány. Powered by TCPDF (www.tcpdf.org)
|
host ::
RSS.