Land cover classification plays a critical role in environmental monitoring, agriculture, urban planning, and disaster management. Traditional methods for land cover classification from satellite imagery are often labor-intensive and prone to errors, highlighting the need for automated and efficient solutions. This project investigates the use of deep learning methods, particularly Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs), for classifying land cover in high-resolution satellite images. The objectives include developing and evaluating deep learning methods to automate the land cover classification task and improve classification accuracy compared to traditional methods. Experimental results demonstrate promising performance, with CNNs achieving an accuracy of 86.12% and ViTs achieving an accuracy of 87.39%. The models exhibit competitive performance in terms of accuracy, efficiency, and scalability, highlighting their utility for any real-world applications in environmental monitoring and geospatial analysis. Future research directions include exploring multi-modal data fusion techniques, advanced attention mechanisms, and integration of domain-specific knowledge to further enhance the capabilities of deep learning methods for classification of land cover.