Comparative Analysis of Machine Learning Models for Satellite Image Classification with Explainable AI Visualization

Satellite image classification is vital for applications like environmental monitoring, urban planning, and resource management. Manual methods are time-consuming and error-prone, highlighting the need for automated solutions. This study explores classifying satellite imagery with four distinct classes using sensor data and Google Map snapshots. Four machine learning models—Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Gradient Boosting Machine (GBM), and Artificial Neural Network (ANN)—were implemented and assessed. The SVM model achieved the highest accuracy at 99%, followed by GBM with 98% accuracy, while KNN attained 94%. Although ANN had the lowest accuracy of 58%, it displayed fewer misclassifications. Explainable AI (XAI) tools like LIME and SHAP were used to interpret model decisions, highlighting how features influenced classifications. The results show machine learning's ability to automate satellite image classification with accuracy and actionable insights.