GREENHOUSES DETECTION USING AN ARTIFICIAL NEURAL NETWORK WITH A VERY HIGH RESOLUTION SATELLITE IMAGE

Detecting and locating greenhouses in south-east of Spain is very important for politicians and other persons who may take decisions about management of natural resources and who must design agricultural development plans. Agriculture is one of more important economic activities in this zone, and till now, development and disposition of new greenhouses was uncontrolled. In this paper, we present a methodology to detect greenhouses from 1.5 m pixel size QuickBird image, based in Artificial Neural Network algorithm. Thanks to the information introduced as training sites, we “teach” to the mathematical model to classify the image considering its radiometric and wavelet texture properties. This assessment is known as training, and the algorithm to obtain it, back-propagation. Classification accuracy was evaluated using multi-source data, comparing results including and no-including wavelet texture analysis. We conclude that some texture analysis can not improve classification accuracy but if one choose correctly parameters and texture model, it can become better. Actually we are working on automatic detection and actualization of greenhouses distributions.

[1]  Jon Atli Benediktsson,et al.  Conjugate-gradient neural networks in classification of multisource and very-high-dimensional remote sensing data , 1993 .

[2]  Soe Win Myint,et al.  A study of lacunarity-based texture analysis approaches to improve urban image classification , 2005, Comput. Environ. Urban Syst..

[3]  P. Swain,et al.  Neural Network Approaches Versus Statistical Methods In Classification Of Multisource Remote Sensing Data , 1990 .

[4]  John R. Jensen,et al.  Introductory Digital Image Processing: A Remote Sensing Perspective , 1986 .

[5]  Giles M. Foody,et al.  Land Cover Classification by an Artificial Neural Network with Ancillary Information , 1995, Int. J. Geogr. Inf. Sci..

[6]  Jon Atli Benediktsson,et al.  Neural Network Approaches Versus Statistical Methods in Classification of Multisource Remote Sensing Data , 1989, 12th Canadian Symposium on Remote Sensing Geoscience and Remote Sensing Symposium,.

[7]  Robert J. Schalkoff,et al.  Pattern recognition - statistical, structural and neural approaches , 1991 .

[8]  Johannes R. Sveinsson,et al.  Feature extraction for multisource data classification with artificial neural networks , 1997 .

[9]  Robert M. Haralick,et al.  Textural Features for Image Classification , 1973, IEEE Trans. Syst. Man Cybern..

[10]  Giles M. Foody,et al.  Using prior knowledge in artificial neural network classification with a minimal training set , 1995 .

[11]  John R. Jensen.  Introductory Digital Image Processing , 2004 .

[12]  W. Hargrove,et al.  Lacunarity analysis: A general technique for the analysis of spatial patterns. , 1996, Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics.

[13]  Martin Brown,et al.  Neurofuzzy adaptive modelling and control , 1994 .

[14]  P. Atkinson,et al.  Introduction Neural networks in remote sensing , 1997 .

[15]  Julius T. Tou,et al.  Pattern Recognition Principles , 1974 .

[16]  Richard P. Lippmann,et al.  An introduction to computing with neural nets , 1987 .

[17]  Lefteri H. Tsoukalas,et al.  Fuzzy and neural approaches in engineering , 1997 .

[18]  R. Lippmann,et al.  An introduction to computing with neural nets , 1987, IEEE ASSP Magazine.

[19]  Richard L. Bankert,et al.  Cloud Classification of AVHRR Imagery in Maritime Regions Using a Probabilistic Neural Network , 1994 .

[20]  S. Cote,et al.  A neural network-based method for tracking features from satellitesensor images , 1995 .