Cloud and Clear Sky Pixel Classification in Ground-Based All-Sky Hemispherical Digital Images*

Cloudiness is the non-predictable factor most affecting the solar radiation reaching a particular location on the Earth surface. Therefore it has great impact on the performance of predictive solar radiation models for that location. This work represents one step towards the improvement of such models by using ground-to-sky hemispherical colour digital images as a means to estimate the fraction of visible sky corresponding to clouds and to clear sky. The general approach, common to many image processing applications, consists in finding one threshold on a given pixel intensity scale that segments the image pixels into clear sky and cloud. In order to allow the evaluation and comparison of image thresholding methods, the pixels of 410 images were manually classified as clear sky or cloud, establishing a reference database. Two well known image thresholding algorithms are tested and a neural network approach is presented. For the latter, a number of statistical measures is extracted from the images constituting a feature space of potential inputs for the neural network. The actual inputs and number of neurons to be employed are selected by means of a multi-objective genetic algorithm.