A new approach to land-based cloud classification

Land-based cloud-classification resolves details which are unavailable in operational satellite imagery. In this study, the emphasis is on single-channel image processing employing convolution masks and statistical measures. Experimental results of classification are examined as a means of deriving simplified cloud amount and class encoding similar to the internationally accepted practice.

[1]  Kenneth A. Buch,et al.  Cloud classification using whole-sky imager data , 1995 .

[2]  Olli Simula,et al.  Experiences from operational cloud classifier based on self-organizing map , 1994, Defense, Security, and Sensing.

[3]  I. Leuthäusser.  Neural network methods , 1991 .

[4]  Olli Simula,et al.  Development of a Cloud Classifier , 1995 .

[5]  Teuvo Kohonen,et al.  The self-organizing map , 1990 .

[6]  Josef Kittler,et al.  Pattern recognition : a statistical approach , 1982 .

[7]  Ari Visa,et al.  Identification of stochastic textures with multiresolution features and self-organizing maps , 1990, [1990] Proceedings. 10th International Conference on Pattern Recognition.

[8]  Charles L. Hull,et al.  Sloan Digital Sky Survey cloud scanner , 1994, Astronomical Telescopes and Instrumentation.

[9]  Olli Simula,et al.  Cloud detection based on texture segmentation by neural network methods , 1991, [Proceedings] 1991 IEEE International Joint Conference on Neural Networks.