The Satellite Segmentation Model project focuses on semantic segmentation of satellite images using deep learning techniques.
All the required setup is done in the notebook itself , just install the libraries as given in jupyter notebook
"satellite_segmentation_full.h5" this is the model you can load it using load_model library
The UNet Model is trained using the model.fit() function with specified batch size, epochs, and validation data.
- Evaluation metrics such as accuracy and Jaccard index are computed during training and saved for analysis.
- Visualizations of input images, ground truth masks, and segmented outputs are generated for qualitative assessment. .