Ground-based Vision Cloud Image Classification based on Extreme Learning Machine
Zhengping Wu, Xian Zhong Xu, Min Xia, Ma Meifang, Lin Li · The Open Cybernetics & Systemics Journal · 2015
Cloud radiation properties and distribution significantly affect the forecasting accuracy, climate monitoring effectiveness and global climate's change.A simple method was proposed to automatically recognize four different sky conditions (cirrus, cumulus, stratus and clear sky) by means of extracting some features from visual images that can be used for training classifier.In this paper, texture features, color features and SIFT features were extracted and extreme learning machine was used for cloud-type classification under different experimental conditions.The experiment results show that the proposed approach using texture features, color features and SIFT features together showed better performance than using these features alone or any two of them together.The accurate identification rate of cirrus, cumulus, stratus and clear sky were 87.67%, 90.75%, 74.50% and 93.63%, respectively with an average of 86.64%.Under the same experimental condition, the proposed method outperformed the artificial neutral network (ANN), k-nearest neighbor (KNN) and support vector machine (SVM).