|
|
Blind Image Deconvolution in STEM mode of Electron Microscope
Valterová, Eva ; Walek, Petr (oponent) ; Potočňák, Tomáš (vedoucí práce)
Blind image deconvolution is method, which restore the true image and point spread function simultaneously. The goal of this paper is to introduce several methods of blind deconvolution and find the optimal method for reconstruction of the true image and point spread function of images from scanning transmission electron microscope. The alternating minimization algorithm is assumed as the most convenient for blind deconvolution problem. Then it is modified and tested. The proposed algorithm properties are tested on artificially degraded data and the real data from scanning transmission electron microscope. The algorithm efficiency is evaluated by several evaluating criteria. The algorithm limitations are determined and its area of use is specified.
|
|
|
Zvyšování kvality videa pomocí konvolučních sítí
Skácel, David ; Špaňhel, Jakub (oponent) ; Hradiš, Michal (vedoucí práce)
Konvoluční neuronové sítě dnes představují v oblasti zpracování obrazu jeden z nejmodernějších přístupů k řešení problémů, jakými jsou například redukce kompresních artefaktů či zvyšování prostorového rozlišení obrazu. Některé výzkumné skupiny již dokazují, že lze tyto sítě adaptovat ke zpracování videa a využít tak přidané informace v čase ke zvětšení prostorového rozlišení videa či dosáhnout lepší úrovně komprese při zachování detailů. Otázkou, zdali je možné využít tento přístup také pro zvýšení časového rozlišení reálného videa, se zabývám v této práci. K tomu využívám konvolučních neuronových sítí, které, jak popisuji, dokáží do jisté míry interpolovat vstupní videosnímky ze skutečných videozáznamů, jsou-li dostatečně kvalitní, a napomoci tak zvýšení snímkové frekvence videa. Dosažené výsledky, ač pozitivní, jsou spíše mezikrokem na cestě za vhodnějším využitím těchto sítí k řešení daného problému.
|
|
|
Blind Image Deconvolution of Electron Microscopy Images
Schlorová, Hana ; Odstrčilík, Jan (oponent) ; Walek, Petr (vedoucí práce)
Blind deconvolution has spread around multiple technical fields in recent years. Problems with computational demands are no more its limitations. Blind deconvolution signal processing techniques are promising solution for enhancement of electron microscope performance. The aim of this work is the problem formulation and proposition of appropriate solution for blind deconvolution of electron microscope images. The final goal is to develop Matlab algorithm correcting aberrations arising from imperfections of image formation and its comparison with built-in Matlab approach implemented in Image Processing Toolbox. Proposed approach is given by regularization techniques of blind deconvolution.
|
|
|
Image Restoration Based on Convolutional Neural Networks
Svoboda, Pavel ; Baláž, Teodor (oponent) ; Sojka, Eduard (oponent) ; Zemčík, Pavel (vedoucí práce)
A merit of this thesis is to introduce a unified image restoration approach based on a convolutional neural network which is to some degree degradation type independent. Convolutional neural network models were trained for two different tasks, a motion deblurring of license plate images and a removal of artifacts related to lossy image compression. The capabilities of such models are studied from two main perspectives. Firstly, how well the model can restore an image compared to the state-of-the-art methods. Secondly, what is the model's ability to handle several ranges of the same degradation type. An idea of the unified end-to-end approach is based on a recent development of neural networks and related deep learning in a field of computer vision. The existing hand-engineered methods of image restoration are often highly specialized for a given degradation type and in fact, define state of the art in several image restoration tasks. The end-to-end approach allows to directly train the required model on specifically corrupted images, and, further, to restore various levels of corruption with a single model. For motion deblurring, the end-to-end mapping model derived from models used in computer vision is deployed. Compression artifacts are restored with similar end-to-end based model further enhanced using specialized objective functions together with a network skip architecture. A direct comparison of the convolutional network based models and engineered methods shows that the data-driven approach provides beyond state-of-the-art results with a high ability to generalize over different levels of degradations. Based on the achieved results, this work presents the convolutional neural network based methods suggesting a possibility having the unified approach used for wide range of image restoration tasks.
|
|
|
Image Restoration Based on Convolutional Neural Networks
Svoboda, Pavel ; Baláž, Teodor (oponent) ; Sojka, Eduard (oponent) ; Zemčík, Pavel (vedoucí práce)
A merit of this thesis is to introduce a unified image restoration approach based on a convolutional neural network which is to some degree degradation type independent. Convolutional neural network models were trained for two different tasks, a motion deblurring of license plate images and a removal of artifacts related to lossy image compression. The capabilities of such models are studied from two main perspectives. Firstly, how well the model can restore an image compared to the state-of-the-art methods. Secondly, what is the model's ability to handle several ranges of the same degradation type. An idea of the unified end-to-end approach is based on a recent development of neural networks and related deep learning in a field of computer vision. The existing hand-engineered methods of image restoration are often highly specialized for a given degradation type and in fact, define state of the art in several image restoration tasks. The end-to-end approach allows to directly train the required model on specifically corrupted images, and, further, to restore various levels of corruption with a single model. For motion deblurring, the end-to-end mapping model derived from models used in computer vision is deployed. Compression artifacts are restored with similar end-to-end based model further enhanced using specialized objective functions together with a network skip architecture. A direct comparison of the convolutional network based models and engineered methods shows that the data-driven approach provides beyond state-of-the-art results with a high ability to generalize over different levels of degradations. Based on the achieved results, this work presents the convolutional neural network based methods suggesting a possibility having the unified approach used for wide range of image restoration tasks.
|
|
|
Blind Image Deconvolution in STEM mode of Electron Microscope
Valterová, Eva ; Walek, Petr (oponent) ; Potočňák, Tomáš (vedoucí práce)
Blind image deconvolution is method, which restore the true image and point spread function simultaneously. The goal of this paper is to introduce several methods of blind deconvolution and find the optimal method for reconstruction of the true image and point spread function of images from scanning transmission electron microscope. The alternating minimization algorithm is assumed as the most convenient for blind deconvolution problem. Then it is modified and tested. The proposed algorithm properties are tested on artificially degraded data and the real data from scanning transmission electron microscope. The algorithm efficiency is evaluated by several evaluating criteria. The algorithm limitations are determined and its area of use is specified.
|
|
|
Zvyšování kvality videa pomocí konvolučních sítí
Skácel, David ; Špaňhel, Jakub (oponent) ; Hradiš, Michal (vedoucí práce)
Konvoluční neuronové sítě dnes představují v oblasti zpracování obrazu jeden z nejmodernějších přístupů k řešení problémů, jakými jsou například redukce kompresních artefaktů či zvyšování prostorového rozlišení obrazu. Některé výzkumné skupiny již dokazují, že lze tyto sítě adaptovat ke zpracování videa a využít tak přidané informace v čase ke zvětšení prostorového rozlišení videa či dosáhnout lepší úrovně komprese při zachování detailů. Otázkou, zdali je možné využít tento přístup také pro zvýšení časového rozlišení reálného videa, se zabývám v této práci. K tomu využívám konvolučních neuronových sítí, které, jak popisuji, dokáží do jisté míry interpolovat vstupní videosnímky ze skutečných videozáznamů, jsou-li dostatečně kvalitní, a napomoci tak zvýšení snímkové frekvence videa. Dosažené výsledky, ač pozitivní, jsou spíše mezikrokem na cestě za vhodnějším využitím těchto sítí k řešení daného problému.
|
|
|
Blind Image Deconvolution of Electron Microscopy Images
Schlorová, Hana ; Odstrčilík, Jan (oponent) ; Walek, Petr (vedoucí práce)
Blind deconvolution has spread around multiple technical fields in recent years. Problems with computational demands are no more its limitations. Blind deconvolution signal processing techniques are promising solution for enhancement of electron microscope performance. The aim of this work is the problem formulation and proposition of appropriate solution for blind deconvolution of electron microscope images. The final goal is to develop Matlab algorithm correcting aberrations arising from imperfections of image formation and its comparison with built-in Matlab approach implemented in Image Processing Toolbox. Proposed approach is given by regularization techniques of blind deconvolution.
|
host ::
RSS.