AUTOMATIC CLASSIFICATION OF VIS-IR METEOSAT IMAGES

This paper analyzes the classification of meteorological images provid ed by geostationary satellites. Classified datasets are used often to gather spatial statistics efficiently from large areas and to simplify the analysis of complex images. The algorithm uses a modified version of k-means standard method to classify the images and statistical dis tance between classes in order to reduce and optimize the final partit ion to be obtained. The method is checked for several meteorological s ituations, both in summer and in winter. The program was written in C and was tested both in UNIX work stations and in personal computers. I n all situations the computer time required to make the classification can be considered acceptable. The methodology presented can be applie d directly to the analysis of meteorological images, providing a tool for interpreting complex scenes. Additionally, minor changes introduce d in the algorithm allow classification of any image provided that two (or more) channels of the same scene are available. Copyright (C) 199 6 Elsevier Science Ltd