Cloud Classification-Based Fine KNN Using Texture Feature and Opponent Color Features

AbstractWeather forecasting and alerts are issued by the meteorological department to alert citizens. Cloud image classification is a challenging task for the research community of meteorology and machine learning. Most of the researches for cloud classification are based on histogram and machine learning using texture features. In this manuscript, the opponent color feature and fine KNN are used to classify the 11 types of cloud images and achieved an accuracy of 99.5% and AUC of 1.0.KeywordsCloud classificationOpponent color featureMeteorologyFine KNN