Журнал

About Editorial Board Information for authors and reviewers Publication ethics Contacts User Account: send / review a manuscript

Multi-layer cloud classification from MODIS data using neural network technology and fuzzy logic approach

V.G. Astafurov^{1, 2} , A.V. Skorokhodov¹

¹ V.E. Zuev Institute of Atmospheric Optics SB RAS, Tomsk, Russia
² Tomsk State University of Control Systems and Radioelectronics, Tomsk, Russia

This paper proposes an automatic layered classification method for multi-layer cloudiness. This method is based on using textural information of MODIS satellite data with a spatial resolution of 250 m and application of artificial neural network technology and fuzzy logic approach. Cloud classification is presented which is based on current meteorological standard and Cloud Atlas. Four methods of statistical approach to texture image description are considered: Gray-Level Co-occurrence Matrix, Gray-Level Difference Vector, Sum and Difference Histogram and One-Dimensional Signal Histogram. We describe the technique for adjusting parameters of a hybrid classifier based on definition of information content of textural features by the Add method. We present results of formation of effective classification features and classification rating achieved in classifying single-layer cloudiness and vertical development clouds. The most frequently occurring combinations were indentified of either of the three cloud layers observed simultaneously over Tomsk territory in the period from 2008 to 2012 (two- and three-layer cloudiness). The multi-layer cloud classification results based on test samples and prospects for further work are discussed. Image classification examples based on full-size MODIS satellite imagery of Tomsk region are demonstrated.

Keywords: information value, classification, neural networks, fuzzy logic, cloudiness, satellite image, textural features, layer

Full text

References:

Astafurov V.G., Skorokhodov A.V., Segmentatsiya sputnikovykh snimkov oblachnosti po teksturnym priznakam na osnove neirosetevykh tekhnologii (Segmentation of cloudiness satellite images by textural parameters based on neural network technologies), Issledovanie Zemli iz kosmosa, 2011, No. 6, pp. 10−20.
Astafurov V.G., Skorohodov A.V. Primenenie nejrosetevyh tehnologij dlja klassifikacii oblachnosti po teksture snimkov MODIS vysokogo razreshenija (The application of neural network technology for the classification of clouds high-resolution MODIS texture images), Issledovanie Zemli iz kosmosa, 2014, No. 5, pp. 39−49.
Bespalov D.P., Devyatkin A.M., Dovgalyuk Yu.A., Kondratyuk V.I., Kuleshov Yu.V., Svetlova T.P., Suvorov S.S., Timofeev V.I., Atlas oblakov (Cloud atlas), Saint-Petersburg: D’ART, 2011, 248 p.
Vyatchenin D.A., Nechetkie metody avtomaticheskoi klassifikatsii (Fuzzy methods of automatic classification), Minsk: UP “Tekhnoprint”, 2004, 219 p.
Volkova E.V. Ocenki parametrov oblachnogo pokrova, osadkov i opasnyh javlenij pogody po dannym radiometra AVHRR c MISZ serii NOAA kruglosutochno v avtomaticheskom rezhime (Automatic estimation of cloud cover and precipitation parameters obtained by AVHRR NOAA for day and night conditions), Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2013, Vol. 10, No. 3, pp. 66−74.
Federal'naya sluzhba po gidrometeorologii i monitoringu okruzhayushchei sredy (Rosgidromet), Kod dlya operativnoi peredachi dannykh prizemnykh meteorologicheskikh nablyudenii s seti stantsii Rosgidrometa (Code for rapid data transfer of surface meteorological observations from a network of Hydromet stations), Moscow: Triada. Ltd, 2013, 79 p.
Zagoruiko N.G., Prikladnye metody analiza dannykh i znanii (Applied methods of data analysis and knowledge), Novosibirsk: IM SO RAN, 1999, 270 p.
Kolodnikova N.V., Obzor teksturnykh priznakov dlya zadach raspoznavaniya obrazov (Overview of textural features for pattern recognition problems), Doklady Tomskogo gosudarstvennogo universiteta sistem upravleniya i radioelektroniki, 2004, Vol. 9, No. 1, pp. 113–124.
Kruglov V.V., Dli M.I., Golunov R.Yu., Nechetkaya logika i iskusstvennye neironnye seti (Fuzzy logic and artificial neural networks), Moscow: Fizmatlit, 2001, 224 p.
Oblaka i oblachnaya atmosfera. Spravochnik (Clouds and cloudy atmosphere. Directory), Leningrad: Gidrometeoizdat, 1989, 647 p.
Osovskii S., Neironnye seti dlya obrabotki informatsii (Neural network for processing information), Moscow: Finansy i statistika, 2002, 344 p.
Bankert R.L., Cloud classification of AVHRR imagery in maritime regions using a probabilistic neural network, Journal of Applied Meteorology, 1994, Vol. 33, No. 8, pp. 909–918.
Bankert R.L., Mitrescu C., Miller S.W., Wade R.H. Comparison of GOES cloud classification algorithms employing explicit and implicit physics, Journal of Applied Meteorology and Climatology, 2009, Vol. 48, pp. 1411–1421.
Baum B.A., Tovinkere V., Titlow J., Welch R.M., Automated cloud classification of global AVHRR data using a fuzzy logic approach, J. Appl. Meteor., 1997, Vol. 36, pp. 1519–1540.
Haralick R.M., Shanmugam K., Dinstein I., Textural features for image classification, IEEE Transactions on Systems, Man and Cybernetics, November 1973, Vol. SMC – 3, No. 6, pp. 610–621.
Weszka J.S., Dyer C.R., Rosenfeld A., A comparative study of texture measures for terrain classification, IEEE Transaction on Systems, Man and Cybernetics, 1976, Vol. SMC – 6, No. 4, pp. 269–285.
Unser M, Sum and difference histograms for texture classification, IEEE Transaction on Systems, Pattern Analysis and Machine Intelligence, 1986, Vol. PAMI – 8, No. 1, pp. 118–125.

создание сайта - TXL