Image Classification of Single Layered Cloud Types 12+

📘 An automatic classification system is presented, which discriminates the different types of single- layered clouds using Principal Component Analysis (PCA) with enhanced accuracy and provides fast processing speed as compared to other techniques. The system is first trained by cloud images. In training phase, system reads major principal features of the different cloud images to produce an image space. In testing phase, a new cloud image can be classified by comparing it with the specified image space using the PCA algorithm. Weather forecasting applications use various pattern recognition techniques to analyze clouds' information and other meteorological parameters. Neural Networks is an often-used methodology for image processing. Some statistical methodologies like FDA, RBFNN and SVM are also being used for image analysis. These methodologies require more training time and have limited accuracy of about 70%. This level of accuracy often degrades classification of clouds, and hence the accuracy of rain and other weather predictions is reduced. PCA algorithm provides a more accurate cloud classification that yield better and concise forecasting of rain.