|
|
Stitching of Retinal Images
Hladyuk, Vadym ; Semerád, Lukáš (referee) ; Drahanský, Martin (advisor)
The purpose of this work is to create a complete picture of the retina, by stitching together a number of partial photos. Since there is no working solution which would be able to capture the entire retina in one picture, this is an important problem to cover. The results will be demonstrated at the end of the text. The problem of stitching partial pictures together was solved by extracting vessels in retinal images, finding key points in images, finding common key points, calculating a transformation matrix and transforming one image into another. After a consultation with an ophthalmologist I was able to define steps which will allow me to further improve the work, which are analyzed in texts. The thesis will provide the reader knowledge about the eye apparatus. It will also introduce field of color models, image formats, algorithms for searching for key points, the transformation of the images themselves and it will also provide a possible way to compose retinal images and also suggest possible improvements.
|
|
|
Traffic Violation Detection on Crossroads
Karpíšek, Miroslav ; Bartl, Vojtěch (referee) ; Špaňhel, Jakub (advisor)
This bachelor thesis presents procedure for the detection of red-light violation. In the theoretical part of the thesis, the current solution aproaches used in image processing are described. The practical part focuses on creation of program for automatic traffic corridors detection, vehicle tracking and the current traffic light state detection. The results obtained by experimenting with the proposed procedure and the possibilities of its further improvement are also discussed.
|
|
|
Panorama Automatically
Motáček, Vladimír ; Španěl, Michal (referee) ; Beran, Vítězslav (advisor)
This paper concerns automatic panoramic image mosaicing. Images can be taken in any direction and in any order. This work uses basic technics such as Harris corner detection, correlation of image patches for finding correspondences and computing homography using RANSAC. The images are mapped to the reference image plane.
|
|
|
Deep Neural Networks for Landmark Detection on 3D Models
Kubík, Tibor ; Kodym, Oldřich (referee) ; Španěl, Michal (advisor)
Detekcia významných bodov je častým krokom pri analýze medicínskych dát. Čoraz bežnejšie sú tieto dáta reprezentované vo forme 3D modelov, príkladom sú povrchové skeny zubného oblúka pacienta. Hlboké neurónové siete sú vhodný spôsob, ako detekovať významné body v obraze. V prípade 3D dát je však toto spracovanie časovo i pamäťovo náročné, čo nevyhovuje požiadavkám kladeným medicínskymi aplikáciami. V tejto práci navrhujem metódu, ktorá tento problém eliminuje a detekuje významné body na povrchu polygonálnych modelov čeľustí. V metóde sú použité rôzne architektúry neurónových sietí, založené na architektúre U-Net. Viacpohľadový prístup presúva spracovanie do 2D, kde navrhnuté architektúry detekujú body regresiou tepelných máp z niekoľkých pohľadov. Pomocou konsezus metódy je následne z týchto pohľadov určená konečná pozícia bodov v 3D priestore. V práci sú predstavené experimenty s dvoma konsenzus metódami: stredná hodnota predikcií a geometrický prístup založený na algoritme RANSAC a metóde najmenších štvorcov. Experimenty ukázali, že varianta kombinujúca Attention U-Net, 100 pohľadov a geometrickú konsenus metódu je schopná detekovať významné body s chybou 1.20 +- 1.81 mm, pričom 94.01% predikcií dosahuje chybu menšiu ako 2 mm.
|
|
|
Basic Shapes Detection in Point Clouds
Eldes, Pavol ; Materna, Zdeněk (referee) ; Španěl, Michal (advisor)
This thesis deals with the topic of simple geometric shapes detection from point-clouds with a special interest in the influence of point normals on speed and quality of produced results. Its main product is an application that demonstrates detection of planes, spheres and cylinders. Detection is done using the RANSAC paradigm that is modified in this thesis to allow it to work with multiple models simultaneously. Another modification presented in this thesis focuses on enforcing stricter candidate shapes selection conditions for detection models.
|
|
|
Detection of Corresponding Points in Images
Komosný, Petr ; Hradiš, Michal (referee) ; Španěl, Michal (advisor)
This thesis is interested in detection of corresponding points in images, which display the same object, eventually some of important elements and synchronizing these images. The aim of this thesis is to find, study and choose suitable algorithm for detecting interesting points in image. This algorithm will be apply at couple of images and in these images will find couples of corresponding points across these images. Functional output of this thesis will be application which will realize choosen interesting points detector, algorithm for finding correspondencies of regions and their synchronizing and joint them to one output image.
|
|
|
Uniform Marker Field on a Cylinder
Kříž, Radim ; Havel, Jiří (referee) ; Herout, Adam (advisor)
This work presents a new extension for Uniform Marker Field, which is able to detect UMF on the cylinder. First part of the text deals with Augmented reality and focuses on systems using markers. It discusses the actual state-of-the-art systems and its possibilities. After that it focuses more deeply on the marker system Uniform marker field and its grayscale variants. Next part of the work describes properties of the cylinder projected in real space. Important properties for detecting are discussed in detail. Then the proposal and description of detection algorithm is presented. Implementation of algorithm is tested and evaluated on the very end of this thesis.
|
|
|
Identification of 3D objects for Robotic Applications
Hujňák, Jaroslav ; Návrat, Aleš (referee) ; Matoušek, Radomil (advisor)
This thesis focuses on robotic 3D vision for application in Bin Picking. The new method based on Conformal Geometric Algebra (CGA) is proposed and tested for identification of spheres in Pointclouds created with 3D scanner. The speed, precision and scalability of this method is compared to traditional descriptors based method. It is proved that CGA maintains the same precision as the traditional method in much shorter time. The CGA based approach seems promising for the use in the future of robotic 3D vision for identification and localization of spheres.
|
|
|
Live Panorama
Jalůvková, Lenka ; Juránková, Markéta (referee) ; Herout, Adam (advisor)
This batchelor`s thesis is focused on creation of live panorama. This is an innovative representation of data from streaming cameras. The input is a video from the streaming camera and the output is also stream - constantly re-drawing panorama according to the current camera position. The resulting solution is based on the SIFT algorithm for finding keypoints in the frames. Furthermore, the RANSAC algorithm is used for calculating homography for stitching images together. Re-drawing the panorama requires weighting of edges. The functionality of the proposed solution was tested on a series of testing video recordings.
|
|
|
Panoramic Photo Creation
Cacek, Pavel ; Čadík, Martin (referee) ; Behúň, Kamil (advisor)
This thesis deals with issues automatic composing panoramic photos from individual photos. Gradually examines the various steps of algorithms and methods used in them, which are used in creating panoramas. It also focuses on the design of the own system based on methods discussed to construct panoramas. This system is implemented using OpenCV library and it is created also a graphical interface using a Qt library. Finally, are in this thesis evaluated outcomes of this designed and implemented system on available datasets.
|
guest ::
