scburning / README.md
jfloresf's picture
Update README.md
0bb7342 verified
---
license: mit
task_categories:
- image-segmentation
language:
- en
tags:
- climate
pretty_name: scburning
size_categories:
- 10K<n<100K
---
## πŸ’Ύ Data
This dataset was used for implementation of an ensemble model for sugar cane burning detection in North and Central Coast of Peru.
For more information, please visit the [project](https://github.com/jfloresf17/scburning)
To download it, use the following commands:
```bash
git clone https://huggingface.co/datasets/jfloresf/scburning
cd scburning
```
Revert the multipart compression by merging and extracting:
```bash
cat database/database.tar.* | tar -xzvf - -i
cat inference/emergencies/emergencies.tar.* | tar -xzvf - -i
cat inference/pilot/pilot.tar.* | tar -xzvf - -i
```
The dataset consists of 1,054 images (πŸ–Ό 512x512 pixels), organized into different folders:
- **Sentinel-2 10 bands** (`S2`)
- **Normalized Burn Ratio** (`nbr`)
- **Burned Area Detection Index** (`badi`)
- **Slope in percentage** (`slope`)
- **Normalized Difference Vegetation Index** (`ndvi`)
- **Normalized Difference Water Index** (`ndwi`)
- **Distance to Crop Cover** (`dlc`)
- **Ground Truth** (`target`)
Additionally, the dataset includes the **emergencies** (`58 images`) and **pilot** (`30 images`) folders for inference purposes with the same structure, except for the **Ground Truth** subfolder.
### Folder Structure
```
database/
β”œβ”€β”€ badi/
β”‚ β”œβ”€β”€ ROI_0000_01_04__..._20201008T082324.tif
| β”œβ”€β”€ ...
β”œβ”€β”€ dlc/
β”‚ β”œβ”€β”€ ROI_0000_01_04__..._20201008T082324.tif
| β”œβ”€β”€ ...
β”œβ”€β”€ nbr/
β”‚ β”œβ”€β”€ ROI_0000_01_04__..._20201008T082324.tif
| β”œβ”€β”€ ...
β”œβ”€β”€ ndvi/
β”‚ β”œβ”€β”€ ROI_0000_01_04__..._20201008T082324.tif
| β”œβ”€β”€ ...
|── ndwi/
| β”œβ”€β”€ ROI_0000_01_04__..._20201008T082324.tif
| β”œβ”€β”€ ...
|── slope/
| β”œβ”€β”€ ROI_0000_01_04__..._20201008T082324.tif
| β”œβ”€β”€ ...
|── S2/
| β”œβ”€β”€ ROI_0000_01_04__..._20201008T082324.tif
| β”œβ”€β”€ ...
|── target/
| β”œβ”€β”€ ROI_0000_01_04__..._20201008T082324.tif
| β”œβ”€β”€ ...
```