Institutional Repository of Vilnius University
Bezje kreivės erdviniam ir laikiniam palydovinių vaizdų žymėjimui /
Abstract
dc:descriptionWith the increasing volume of data and the widespread application of deep learning techniques across various domains, such as agriculture, where the goal is to predict and manage crop variability; geology, where satellite images are utilized to gather information on different terrain surfaces, their changes, or water permeability; and epidemiology, where satellite images can identify the spread of diseases or infections, there is a growing need for high-quality datasets for research and model training. This need is highlighted in scientific publications addressing segmentation tasks related to satellite image segmentation. In this work, to mitigate the impact of incomplete datasets on machine learning algorithms, we review essential methods for improving segmentation models, identify data annotation techniques, and introduce and examine a new method for interpolating satellite image sets. This method involves averaging images and masks over time using a Bezier curve, and its effectiveness is evaluated through a comparative study.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Sinkevičiūtė, Ema,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD210647979&prefLang=en_US