Satellite Image Classification Using Convolutional Neural Networks

Image classification is crucial in remote sensing, image analysis, and pattern recognition. Object detection and classification in satellite images are crucial for applications like marine monitoring, land-use planning, ecological research, and military operations. Satellite images provide valuable spatial and temporal data, helping address real-world challenges. However, issues like data quality, availability, and distribution complicate image analysis. Deep learning (DL) algorithms have become popular for their effectiveness in classifying satellite images, sup-porting tasks such as disaster management, land-use planning, and natural resource management. This study proposed a CNN model with batch normalization for satellite image classification. The dataset includes images categorized into cloudy regions, deserts, water bodies, and green spaces. CNN excels at extracting features like edges and textures, while Batch Normalization (BN) stabilizes training by normalizing inputs and reducing covariate shifts. Together, CNN and BN improve the efficiency and accuracy of the model. The proposed model successfully classifies satellite images, outperforming existing methods with an overall accuracy of 91%, demonstrating its superior remote sensing imaae analysis capability.