Characterisation of clouds and their heights by texture analysis of multi-spectral stereo images

Non-linear filtering and representation of singular values by a Lebesgue measure and Lebesgue integral create the values to characterise different properties. Fuzzy measures and fuzzy functions represent these different types of properties, stochastic as well as non-stochastic. This is the basis for their fusion by fuzzy integrals for the characterisation of clouds and estimation of their boundary. The boundaries of the clouds are described by parameter functions obtained by fuzzy curve analysis in different directions. Such functions characterising the boundary are generated for different spectral images. This is used for characterising points for matching and finding similarity of parts of the clouds for stereo matching and estimation of the heights of clouds