Multi-Model AI of Satellite Image Processing for Enhanced Urban Development Analysis in Bangladesh

The rapid growth of urbanization in Bangladesh impacts both landscapes and environments, creating challenges for policymakers and urban planners. Consequently, advanced monitoring methods are necessary to address these challenges. This research addresses the need for a detailed satellite imagery dataset by leveraging AI and deep learning for image processing, including annotation. We created a dataset by collecting and annotating satellite images from various regions of Bangladesh, identifying buildings, trees, water sources, and farmland. This dataset enables the analysis of landscape changes, urban growth, environmental degradation, and land encroachments over the past decade. Our system integrates four state-of-the-art deep learning models to ensure high accuracy and reliable object detection through cross-verification. Despite challenges with data imbalance, our approach shows promise in enhancing AI performance for satellite imagery analysis. Future work will focus on expanding the dataset, addressing data imbalance, and incorporating additional models to further improve accuracy. This research provides valuable insights for urban planning, environmental conservation, and national security.