|
|
Blind Image Deconvolution in STEM mode of Electron Microscope
Valterová, Eva ; Walek, Petr (referee) ; Potočňák, Tomáš (advisor)
Slepá dekonvoluce je metoda, při které je rozptylová funkce a skutečný obraz rekonstruován zároveň. Cílem této práce je představit různé metody slepé dekonvoluce a najít optimální metodu rekonstrukce původního obrazu a rozptylové funkce. Jako nejvhodnější metoda slepé dekonvoluce byl zvolen algoritmus střídavé minimalizace, který byl upraven a testován. Vlastnosti navrženého algoritmu byly testovány na uměle degradovaných datech a na reálných datech pořízených skenovacím transmisním elektronovým mikroskopem. Účinnost algoritmu byla hodnocena hned několika hodnotícími kritérii. Byla zjištěna omezení algoritmu a tím specifikováno jeho využití.
|
|
|
Detection of delay between arterial and venous pulse signal from videoophthalmoscope
Pejskar, Michael ; Mézl, Martin (referee) ; Valterová, Eva (advisor)
The goal of this bachelor´s work is to measure a time difference between arterial and venous pulses on a retina from video sequences obtained from a binocular videoopthalmoscope. Theoretical part contains a description of the eye and its structures, description of the imaging devices used to display ocular background and research of methods used to measure and calculate the delay between pulses. Practical part describes the processing and adjustments of the videosequences and the measurement of time differences using foot-to-foot and phase difference method.
|
|
|
Radiation damage of samples in low-voltage transmission electron microscopy
Vykydal, Václav ; Valterová, Eva (referee) ; Kolář, Radim (advisor)
This thesis is focused on Radiation damage of samples in low-voltage transmission electron microscopy. In thesis is general description of transmission electron microscopy and their important parts. There is described how samples are damaged by primary electron beam and how samples for transmission electron microscopy are prepared and which degradation occurs from the point of view materials of sample and time dose of electrons from primary electron beam.
|
|
|
Processing of high-resolution retinal images
Vraňáková, Sofia ; Chmelík, Jiří (referee) ; Valterová, Eva (advisor)
Diplomová práca je zameraná na spracovávanie obrazov sietnice s vysokým rozlíšením. Cieľom práce je zlepšiť výslednú kvalitu výsledných snímkov sietnice získaných zo sekvencie snímkov nižšej kvality. Jednotlivé snímky sú najskôr spracované pomocou bilaterálnej filtrácie a zlepšenia kontrastu. v ďalšom kroku sú odstránené rozmazané snímky a snímky zobrazujúce iné časti sietnice. Posun medzi jednotlivými snímkami v sekvencii sa odhaduje pomocou fázovej korelácie, a tieto obrazy sú potom fúzované do výsledného snímku s vysokým rozlíšením pomocou priemerovania a využitia superrozlišovacej techniky, presnejšie regularizácie pomocou bilaterálneho celkového rozptylu. Výsledné mediánové hodnoty skóre kvality získaných obrazov sú PIQUE 0.2600, NIQE 0.0701, a BRISQUE 0.3936 pre techniku priemerovania, a PIQUE 0.1063, NIQE 0.0507, and BRISQUE 0.1570 pre superrozlišovaciu techniku.
|
|
|
Plethysmography for retinal video processing
Plavcová, Daniela ; Valterová, Eva (referee) ; Kolář, Radim (advisor)
Sight is one of the five human senses and is very important for exploring near and far surroundings. However, this is prevented by various eye defects, which may be related to the vascular system of the retina and the pulsations that are observable on it. This work deals with the study of the relationship between the mentioned pulsations and glaucoma. To search for connections between these two aspects, video recordings from the retina are used, the pulsation of which is further worked in the form of a 1D signal, especially in the spectral region.
|
|
|
Blood vessel segmentation in high resolution retinal images
Svobodová, Sabina ; Odstrčilík, Jan (referee) ; Valterová, Eva (advisor)
This thesis focuses on implementation of an algorithm for retinal vessel segmentation in high resolution retinal images.A neural network with two hidden layers was used as the method. A total of 7 features were obtained from matched filtering based on vessel thickness, texture analysis and individual pixels brightness. Within the thesis, the whole database was manually annotated for the implementation of the algorithm and the results. The achieved mean sensitivity reached 80%, specificity 70% and Dice coefficient is 59%.
|
|
|
Spatial registration of partially overlapping retinal images
Žiak, Milan ; Harabiš, Vratislav (referee) ; Valterová, Eva (advisor)
Adaptive optics give us opportunity to capture high resolution images of retina, even we are able to detect cones. Spatial registration of retina images has to be done due to scanning process of capturing, because current opthalmoscops with adaptive optics aren't able to capture whole area of retina. Diploma thesis demonstrates basic princip of automatic spatial registration of overlaping images. It uses constellation features for matching some of the detected cones. K-means combinate wiht RANSAC is used for outlier detection. Inliers matching pairs are then used to define a affine transformation matrix. This method was aplied on dataset retina images from AO-FIO modality and evaluation of quality was computed according to normalized mutual information. Result shows that this metod is suitable for registration AO-FIO images with "cell-to-cell" accuracy.
|
|
|
Spatial registration of partially overlapping retinal images
Žiak, Milan ; Harabiš, Vratislav (referee) ; Valterová, Eva (advisor)
Adaptive optics give us opportunity to capture high resolution images of retina, even we are able to detect cones. Spatial registration of retina images has to be done due to scanning process of capturing, because current opthalmoscops with adaptive optics aren't able to capture whole area of retina. Diploma thesis demonstrates basic princip of automatic spatial registration of overlaping images. It uses constellation features for matching some of the detected cones. K-means combinate wiht RANSAC is used for outlier detection. Inliers matching pairs are then used to define a affine transformation matrix. This method was aplied on dataset retina images from AO-FIO modality and evaluation of quality was computed according to normalized mutual information. Result shows that this metod is suitable for registration AO-FIO images with "cell-to-cell" accuracy.
|
|
|
Automatic Adaptive Optics Retinal Images Montaging
Valterova, Eva
The adaptive optics images capture limited field of view. Nevertheless, the quantitativeretina assessment requires analysis over extended areas captured by images on various retina positions.The fully automated method for image registration and montaging is presented. The methodutilizes scale invariant feature transform (SIFT) for feature extraction from preprocessed images.The method is tested on 200 images of normal healthy patients. The montages of 20 eyes were createdand evaluated by normalized mutual information metrics and the results showed high alignmentaccuracy.
|
|
|
Automated Fovea Center Estimation In Adaptive Optics Images
Valterova, Eva
The adaptive optics (AO) is an insightful tool with high potential. Since the first demonstration in ophthalmology has undergone immense growth in its diagnostic application, mainly because the AO enables to capture retina in vivo with resolution on the cellular level. In the center of the retina is macula, containing the fovea. The fovea detection is possible with utilizing the knowledge of its characteristical properties. We developed and applied two methods for fovea center estimation. The first method is based on image maximum detection. The second is based on the 2D Gaussian curve fitting. The first method has proved better results in comparison with manual grading. The measured difference was 32_x0006_27 pixels. The second approach has in comparison with manual grading slightly higher difference equal to 108_x0006_64 pixels. Both of the methods estimated the foveal center within the expected foveal area.
|
guest ::
RSS feed.