|
1 |
| -## D-Fire: an image dataset for fire and smoke detection |
| 1 | +# D-Fire: an image dataset for fire and smoke detection |
2 | 2 |
|
3 | 3 | **Authors:** Researchers from Gaia, solutions on demand ([GAIA](https://www.gaiasd.com/))
|
4 | 4 |
|
5 |
| -### About |
| 5 | +## About |
6 | 6 |
|
7 | 7 | D-Fire is an image dataset of fire and smoke occurrences designed for machine learning and object detection algorithms with more than 20,000 images.
|
8 | 8 |
|
| 9 | +<div align="center"> |
9 | 10 | <table>
|
10 |
| -<tr><th>Number of images</th><th>Number of bounding boxes</th></tr> |
11 |
| -<tr><td> |
| 11 | + <tr> |
| 12 | + <th>Number of images</th> |
| 13 | + <th>Number of bounding boxes</th> |
| 14 | + </tr> |
| 15 | + |
| 16 | + <tr><td> |
12 | 17 |
|
13 |
| -| Category | # Images | |
14 |
| -| ------------- | ------------- | |
15 |
| -| Only fire | 1,164 | |
16 |
| -| Only smoke | 5,867 | |
17 |
| -| Fire and smoke | 4,658 | |
18 |
| -| None | 9,838 | |
| 18 | + | Category | # Images | |
| 19 | + | ------------- | ------------- | |
| 20 | + | Only fire | 1,164 | |
| 21 | + | Only smoke | 5,867 | |
| 22 | + | Fire and smoke | 4,658 | |
| 23 | + | None | 9,838 | |
19 | 24 |
|
20 |
| -</td><td> |
| 25 | + </td><td> |
21 | 26 |
|
22 |
| -| Class | # Bounding boxes | |
23 |
| -| ------------- | ------------- | |
24 |
| -| Fire | 14,692 | |
25 |
| -| Smoke | 11,865 | |
| 27 | + | Class | # Bounding boxes | |
| 28 | + | ------------- | ------------- | |
| 29 | + | Fire | 14,692 | |
| 30 | + | Smoke | 11,865 | |
26 | 31 |
|
27 |
| -</td></tr> </table> |
| 32 | + </td></tr> |
| 33 | +</table> |
| 34 | +</div> |
28 | 35 |
|
29 |
| -All images were annotated according to the YOLO format (normalized coordinates between 0 and 1). |
| 36 | +All images were annotated according to the YOLO format (normalized coordinates between 0 and 1). |
| 37 | +However, we provide the yolo2pixel function that converts coordinates in YOLO format to coordinates in pixels. |
30 | 38 |
|
31 | 39 | ***
|
32 | 40 |
|
33 |
| -### Download |
| 41 | +## Some examples of labeled images |
| 42 | + |
| 43 | +<p align = "center"> |
| 44 | +<img src="https://drive.google.com/uc?export=view&id=1TfxQUZd89zU7xYG3y0enivXk3rqI6sIj" width="auto" height="180"/> |
| 45 | +<img src="https://drive.google.com/uc?export=view&id=1OywHP44jNIy0uoA6O63VZPB1_Hi-sejK" width="auto" height="180"/> |
| 46 | +<img src="https://drive.google.com/uc?export=view&id=1mIvClp9fLnKURYrRoNab8b96FZCm3Zje" width="auto" height="180"/> |
| 47 | +</p> |
| 48 | + |
| 49 | +## Download |
34 | 50 |
|
35 | 51 | * [Download D-Fire Dataset (only images and labels)](https://drive.google.com/drive/folders/1DWgsQLVgkkLM8m-VcugHNpD5WYDbjYp5?usp=sharing).
|
36 | 52 | * [Download some surveillance videos](https://drive.google.com/drive/folders/1P5TNDP7ZrWpIZ4v_Aav5hf3S9UII2ZKA?usp=sharing).
|
37 | 53 | * If you want more surveillance videos, please request your registration on our environmental monitoring website ["Apaga o Fogo!" (Put out the Fire!)](https://apagaofogo.eco.br/).
|
38 | 54 | * Data splitting is also available [here](https://drive.google.com/drive/folders/1Np_FC3MuuFJgV-z0FmZwS9YzsTKdyRGJ?usp=sharing).
|
39 | 55 |
|
40 |
| -### Citation |
| 56 | +## Citation |
41 | 57 |
|
42 | 58 | Please cite our paper in your publications if you use this dataset.
|
43 | 59 |
|
|
0 commit comments