Institutional Repository of Vilnius University
Neuroninių tinklų taikymas mikroskopinių vaizdų segmentavimui /
Abstract
dc:descriptionThis bachelor's thesis examines the segmentation of microplastics on a nonhomogeneous background. The aim of the study is to propose and investigate a neural network suitable for microplastics segmentation on a nonhomogeneous background. To achieve this goal, a program using the UNet model was developed, which was trained with a dataset of microscopic images of microplastics collected by the physical and technology science center (FTMC), data augmentation was performed, and hyperparameters were optimized. The study results showed that UNet is suitable for microplastics segmentation, but due to the small size of the dataset, the segmentation results are not ideal. To more accurately assess model performance, it is recommended to conduct future studies using more correct, larger and more uniform datasets.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Junevičius, Vilius,
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:ELABAETD210645731&prefLang=en_US