|
|
--- |
|
|
license: mit |
|
|
--- |
|
|
|
|
|
# Satellite Segmentation Model |
|
|
|
|
|
## Description |
|
|
This project contains a deep learning model for satellite image segmentation. The model is trained to classify different classes such as buildings, land, roads, vegetation, water, and unlabeled areas in satellite images. |
|
|
|
|
|
## How to Use |
|
|
To use the model, follow these steps: |
|
|
1. Install the required dependencies by running `pip install -r requirements.txt`. |
|
|
2. Load the trained model (`satellite_segmentation_full.h5`) using your preferred deep learning framework or library. |
|
|
3. Preprocess your input satellite image data according to the model's input requirements (e.g., resize, normalize). |
|
|
4. Feed the preprocessed images to the model for segmentation. |
|
|
5. Obtain the segmented output masks corresponding to different classes. |
|
|
|
|
|
### Input Requirements |
|
|
- Input images should be in RGB format. |
|
|
- Images should be preprocessed to match the model's input size and normalization. |
|
|
|
|
|
### Expected Output Format |
|
|
- Segmented masks representing different classes (e.g., Building, Land, Road, Vegetation, Water, Unlabeled) in the input image. |
|
|
|
|
|
## Dependencies |
|
|
- Python 3.8 |
|
|
- TensorFlow |
|
|
- NumPy |
|
|
- OpenCV |
|
|
- PIL (Python Imaging Library) |
|
|
- Matplotlib |
|
|
- scikit-learn |
|
|
- Other dependencies as specified in `requirements.txt`. |
|
|
|
|
|
## Additional Setup Instructions |
|
|
- For detailed installation and usage instructions, refer to the documentation in the repository or project wiki. |
|
|
- Make sure to have GPU support if training or inference requires significant computational resources. |
|
|
|
|
|
## License |
|
|
This project is licensed under the [MIT License](LICENSE). |
|
|
|
