Identifying urban regions in Very High-Resolution (VHR) satellite images is a critical task for urban planning, environmental monitoring, and disaster management. This study explores the use of computational neural networks (CNNs) for automated detection and classification of urban areas in VHR satellite imagery. Leveraging the deep learning capabilities of CNNs, we propose a model that efficiently captures complex spatial patterns and textures characteristic of urban landscapes. Our methodology involves training the CNN on a diverse dataset of annotated satellite images, ensuring robust performance across various urban environments. The proposed approach demonstrates high accuracy and computational efficiency, significantly improving upon traditional methods. Results in this research show that methods employing CNN have an ability to increase the automation and precision of urban region detection from satellite imagery.