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
    │ ......................

Segmentation_water_bodies_Peru

Dataset Perusat--- HR Dataset Sentinel_l--- LR

Run_HR:

1. python train_HR.py
2. python plotting.py (need path roots)

Run_LR:

1. python train_LR.py
2. python plotting.py (need path roots)

Model Combined Parallel:

1. python train_paral.py
2. python plotting.py (need path roots)