Your image segmentation model, based on ResNet18, ResNet50 and ResNet101 can be used for various applications in remote sensing, urban planning, and environmental monitoring. By analyzing satellite or aerial images, the model can accurately classify and segment different land cover types, such as vegetation, water bodies, and urban areas. This is particularly useful in detecting urban heat islands (UHI), where temperature variations across different land types can be studied to optimize city planning and reduce environmental impact. Additionally, it can assist in disaster management by identifying affected areas in post-disaster scenarios. The model can be deployed in a Streamlit web app, allowing users to upload images and receive segmented outputs in real time. By integrating with Hugging Face Hub, the model remains easily accessible and can be updated for improved performance. This makes it a powerful tool for researchers, urban developers, and environmentalists seeking data-driven insights from satellite imagery.
