International Journal of Engineering and Technology

Weather forecasting needs multiple features to predict in that cloud classification is the one of the feature. Cloud Classification system through satellite images which are taken from multiple bands. The classification is done by combining the Artificial Neural Network (ANN) and Genetic Algorithm (GA). The spectral ratio value of the image is given as input because it is more efficient to detect the spectral properties of cloud and it enhances the image pixel by pixel. Non convex decision handling and misclassifications of classes were predicts using textural features in the Artificial Neural Network. The optimal weight value is finds using Genetic Algorithm that’s combining ANN and GA increases the efficiency of classification and weather prediction. The cloud is separated as eight different classes and the weather temperature is defined using this classification. The combined algorithm of ANN and GA increase the accuracy percentage up to 3% then separate classification algorithm ANN. The overall accuracy of this algorithm is 87.5% for cloud classifications.