Application of Semantic Segmentation with Few Labels in the Detection of Water Bodies from Perusat-1 Satellite’s Images
Dataset: Perusat
First, we explore the application of a U-Net model and then a transfer knowledge-based model (train_paral.py). How to run
The dataset is organized in the folloing way:: ::
├── data_HR
│ ├── test
│ ├── images
│ └── masks
│ └── val
│ ├── images
│ └── masks
│ └── train
│ ├── images
│ └── masks
├── data_LR
│ ├── test
│ ├── images
│ └── masks
│ └── val
│ ├── images
│ └── masks
│ └── train
│ ├── images
│ └── masks
├── logs_LR
│ ├── mapping
│ ├──
├── predictions
├── history
│ ......................
Dataset Perusat--- HR Dataset Sentinel_l--- LR
Run_HR:
- python train_HR.py
- python plotting.py (need path roots)
Run_LR:
- python train_LR.py
- python plotting.py (need path roots)
Model Combined Parallel:
- python train_paral.py
- python plotting.py (need path roots)