| Title |
Automatinė aerofotografinių nuotraukų analizė taikant giliojo mokymosi metodus |
| Translation of Title |
Automatic analysis of aerophotography images using deep learning methods. |
| Authors |
Vyšniauskas, Vykintas |
| Full Text |
|
| Pages |
102 |
| Keywords [eng] |
deep learning ; convolutional neural networks ; computer vision ; remote sensing ; image segmentation |
| Abstract [eng] |
This work reviews capabilities of deep learning-based methods for automatic analysis of aerial images. To reach the goal, the following tasks were completed: analytical review of methods for automatic analysis of aerial images. Selection of viable deep learning methods for automatic aerial images analysis. Experimental study and practical implementation of state-of-the-art segmentation models for aerial images segmentation. Due to complexity of aerial images, diversity of objects and other challenges, current methods of artificial intelligence and machine learning lack precision and therefore human intelligence still needs to be used in remote sensing applications. |
| Dissertation Institution |
Vilniaus Gedimino technikos universitetas. |
| Type |
Master thesis |
| Language |
Lithuanian |
| Publication date |
2021 |
