Institutional Repository of Vilnius University
Lentelės langelių aptikimas pasitelkiant kompiuterinę regą ir neuroninius tinklus /
Abstract
dc:descriptionIn these times importance of computer vision and neural networks is increasing. Digitization and accessibility features for people with disabilities becomes more and more actual, so it is important to move as much data to digital storage. This paper compares computer vision algorithms of SUSAN, Forstner and Harris and also neural network architectures of U-Net, Inception and ResNet. After analysis of these algorithms and networks there is created data set and chosen neural network architecture to train. Trained model is able to detect cell's corners. Results of this model can be used to improve data serialization algorithms to read data from tables.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Šideikis, Justinas,
- Contributors dc:contributor
-
- Valaitis, Vytautas
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:ELABAETD81705016&prefLang=en_US