Derivation of Land Cover Information from Remotely Sensed Data using Multiple Classifiers

Content Provider	Semantic Scholar
Author	Carlos, A. Cardeño C. Vieira, Ricardo Parente Garcia Paul, Catherine J. Morrison Mather
Copyright Year	2001
Abstract	Due to the growing volume of image data from planned and existing sensors, new data-processing techniques are required to allow the information to be processed promptly and accurately. Although the range of image processing techniques has greatly expanded in recent years, from classical statistical approaches to neural network methods, there is no single classification algorithm capable of deriving generic products from remotely sensed data that can be used with confidence. The performance of these algorithms is strongly dependent upon the selected data set and on the efforts devoted to the design phase. In this paper, we report a more systematic investigation into the problem of combining multiple classifiers in the context of land cover mapping using remotely sensed data. Four approaches are proposed, based on voting principles, Bayesian formalism, evidential reasoning, and artificial neural networks. Preliminary results indicate that improvements can be obtained in difficult pattern recognition problems by combining or integrating the outputs of complementary multiple classifiers.
File Format	PDF HTM / HTML
Alternate Webpage(s)	http://iris.sid.inpe.br:1908/col/dpi.inpe.br/lise/2001/09.20.17.58/doc/1353.1365.011.pdf
Alternate Webpage(s)	http://marte.sid.inpe.br/col/dpi.inpe.br/lise/2001/09.20.17.58/doc/1353.1365.011.pdf
Language	English
Access Restriction	Open
Content Type	Text
Resource Type	Article