Weather Classification System using Deep Learning Technology for National Astronomical Research Institute of Thailand SkyCamera

Abstract

The Weather Classification System Using Deep Learning Technology for Sky Cameras of the National Astronomical Research Institute (Public Organization) The objective is to use continuous sky photo data stored in a database to train and develop an artificial intelligence system for weather classification. This aims to increase the accuracy in displaying sky statuses, including clear sky, cloudy, overcast, and rainy conditions. By using deep learning algorithms, the system achieves an accuracy rate of 96.67% through image classification with convolutional neural networks. These networks are based on artificial neural networks that can effectively learn and analyze complex data. The system can process and classify sky conditions in real-time based on specified locations and times. Additionally, it can store data to display through graphs, allowing users to know the sky conditions at different times. The system also provides the capability to retrieve historical data by date, time, and location as needed by the users. This is highly beneficial for studying and analyzing weather conditions and astronomical phenomena.