Sky Classification Using CNNs for Meteorological Condition Detection

Sky image classification under different meteorological conditions is crucial for numerous applications, including environmental monitoring and weather forecasting. In this work, we propose a Convolutional Neural Network (CNN)-based approach to classify sky images into four categories: 'Day-clear', 'Day-cloudy', 'Nigh-clear', and 'Night-cloudy'. Several deep learning architectures were explored, and experimental results show that YOLOv11 achieved the best performance, with an accuracy exceeding 99%. Furthermore, potential future developments are discussed, such as the inclusion of rain as an additional meteorological condition to be classified. This study provides a solid foundation for improving image recognition technologies in meteorological contexts, with potential applications across various technological and scientific fields.