AI-Assisted Manual Segmentation Web Application for Geospatial Satellite and Imagery Data

In recent years, the need for efficient geospatial image editing and analysis tools has increased significantly due to the rapid growth of satellite and aerial imagery data. An AI-Assisted Manual Segmentation Web Application, a ReactJs web application was designed for editing GeoTiffs and generating masks using advanced algorithms such as K-means and U-Net. The application offers a comprehensive suite of editing tools, including pencil, eraser, undo, redo, zoom, magic wand, import mask, export mask, and layers management. These features allow users to manipulate geospatial imagery data effectively and intuitively. Furthermore, our application employs a Flask server to facilitate the generation of K-means or U-Net masks, depending on the user's choice. Upon pressing the corresponding button, the server processes the image and sends the generated mask back to the ReactJs web application. The application seamlessly integrates the generated mask as a new layer, enabling users to further refine their edits and analysis. By combining advanced AI algorithms with a user-friendly interface, our AI-Assisted Manual Segmentation Web Application aims to achieve the goal of improving the current methodology of analysing aerial imagery.