magistrsko delo
Aleš Obal (Author), Dušan Gleich (Mentor)

Abstract

V magistrskem delu je opisan postopek kategorizacije in detekcije sprememb v radarskih slikah SAR z uporabo konvolucijske nevronske mreže, tako imenovane »Deep Learning«. V ta namen je bil izdelan program v Matlab programskem okolju, ki je sposoben izrezati SAR radarsko sliko na poljubne manjše dele, ustvariti konvolucijsko nevronsko mrežo, jo naučiti ter uporabiti pri klasifikaciji in detekciji predelov površja na Zemlji. Programu je dodana tudi opcija uporabe grafično procesne enote »GPU« za pohitritev učenja konvolucijskih nevronskih mrež.

Keywords

kategorizacija slik;detekcija sprememb;globoke nevronske mreže;paralelno procesiranje;SAR;magistrske naloge;

Data

Language: Slovenian
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher: A. Obal
UDC: 004.032.26:550.837.7(043.2)
COBISS: 21209110 Link will open in a new window
Views: 1081
Downloads: 164
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Categorization and change detection using synthetic aperture radar data with deep learning algorithm
Secondary abstract: The master's thesis describes the process of categorization and detection of changes in SAR radar images using a convolutional neural network, the so-called "Deep Learning." A program was developed In the Matlab program environment for this purpose. The program is capable of cutting the SAR radar image into smaller thumbnails, creating convolution neural network for learning and recognition of Earth's surface parts and changes on it. The program also contains an option to use the graphics processing unit »GPU« for speeding up the learning process of the convolutional neural network.
Secondary keywords: image categorization;change detection in image;deep neural network;paraller processing;
URN: URN:SI:UM:
Type (COBISS): Master's thesis/paper
Thesis comment: Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Elektrotehnika
Pages: VIII, 72 str.
ID: 10895506