부산대학교 도서관

Weather forecasting is an important task as it helps people to manage their work. Cloud movement is a major factor for the variation of the weather. Along with the other factors, cloud movements can be used to predict the weather changes. This paper propose a method to analyze the cloud movements and calculate the motion vectors using Lucas Kanade method. Cloud velocity vectors can be used with cloud types, wind and thermal data to predict the cloud patterns which helps to forecast weather changes. Proposed method has implemented on a software platform and experiments has been carried out to validate the system.