|
|
Image classification using deep learning
Hřebíček, Zdeněk ; Přinosil, Jiří (referee) ; Mašek, Jan (advisor)
This thesis deals with image object detection and its classification into classes. Classification is provided by models of framework for deep learning BVLC/Caffe. Object detection is provided by AlpacaDB/selectivesearch and belltailjp/selective_search_py algorithms. One of results of this thesis is modification and usage of deep convolutional neural network AlexNet in BVLC/Caffe framework. This model was trained with precision 51,75% for classification into 1 000 classes. Then it was modified and trained for classification into 20 classes with precision 75.50%. Contribution of this thesis is implementation of graphical interface for object detction and their classification into classes, which is implemented as aplication based on web server in Python language. Aplication integrates object detection algorithms mentioned abowe with classification with help of BVLC/Caffe. Resulting aplication can be used for both object detection (and classification) and for fast verification of any classification model of BVLC/Caffe. This aplication was published on server GitHub under license Apache 2.0 so it can be further implemented and used.
|
|
|
Parallel Programming in Rust Language
Šlampa, Ondřej ; Bařina, David (referee) ; Kobrtek, Jozef (advisor)
Topic of this thesis is parallelization in Rust. Aim of this thesis is to compare performance and usability of Rust language with already used alternative - OpenMP. Computation of n-dimensional comvolution was used for benchmark. In conclusion there is evaluation of results and suggestions for their future use.
|
|
|
Quality assessment of retinal images
Tvarůžek, Marek ; Jiřík, Radovan (referee) ; Kolář, Radim (advisor)
This thesis deals with the properties of retinal images, which are taken with a digital fundus camera. The basic image processing operations are described and applied on biomedical images. Based on these operations, the algorithm for image quality assessment is designed and implemented in Matlab. The quality metrics quantify the focusing. The proposed methods are tested on two datasets and the results are presented and discussed.
|
|
|
3D Volume Rendering Data Visualization
Varga, Marek ; Herout, Adam (referee) ; Kršek, Přemysl (advisor)
This thesis deals with using theoretical knowledge of computer graphic and vector mathematics for volume data construction and vizualization. This work provides more methods volume data vizualization and optimization. Algorithms and mathematical principles are necessary for data obtaining, correct interpretation, filtration and error correction and vizualization. This work describes possibilities of method voxel splatting. Model which implements this method is a part of this work and it contains some techniques for drawing process which include skipping empty sections and effective addressing of properties of volume data. System also provides interactive rendering and volume data manipulation.
|
|
|
Time-Gathering Video
Kárníková, Pavlína ; Sumec, Stanislav (referee) ; Herout, Adam (advisor)
This bachelor essay describes problems of creating time-gathering video, image processing and the theory of photography. Purpose is created program, which can modify photos and after that create video. Program is implemented in the C and C++ language and uses the OpenCV library. Application can run in the operation system Windows.
|
|
|
Techniques Used for Image Smoothing, Blurring and Sharpening
Kubínek, Jiří ; Šilhavá, Jana (referee) ; Venera, Jiří (advisor)
This work is dedicated to methods used for digital image editing. It defines fundamental terms for this work as color space or noise. Above all, it analyses methods allowing image sharpening and blurring. It describes some of the most known algorithms from the theoretical point of view, but also introduces their implementation in C programming language. There are compared according to time complexity. The purpose of this work is to introduce digital image filtering and demonstrate elementary procedures used for their implementation.
|
|
|
Longtime Video
Macháček, Martin ; Beran, Vítězslav (referee) ; Juránková, Markéta (advisor)
This thesis describes methods used for making time-lapse video from a set of photographs and tools, which make it easier. It suggests and evaluates different approaches useful for elimination of frames, which differ from the rest of the set, application of digital image filters and other useful features. Practical part of this work is a computer program implementing these methods and features.
|
|
|
Canny's Operator and Other Useful Edge Detectors
Janda, Miloš ; Juránek, Roman (referee) ; Venera, Jiří (advisor)
This work introduces main approaches for digital image processing and defines fundamental terms for successful understanding. Main aim is description of several suitable methods used in digital image pre-processing, methods for edge detection and consequent post-processing of these. The final goal of this work is effective implementation and complex comparison of methods for edge detection.
|
|
| |
|
|
GPU Image Processing Library
Čermák, Michal ; Španěl, Michal (referee) ; Smrž, Pavel (advisor)
This work is concerned with architecture of recent Nvidia graphics cards and application programming interface CUDA. That is used to create accelerated image processing library. It place emphasis on testing performance gain compassion with high optimized and used OpenCv library.
|
guest ::
