Relevance of airborne lidar and multispectral image data for
urban scene classification using Random Forests, ISPRS
Journal of Photogrammetry and Remote Sensing, 66(1): 56-66.
Haala, N. and Walter, V., 1999. Automatic classification of
urban environments for database revision using lidar and color
aerial imagery, International Archives of Photogrammetry and
Remote Sensing, Vol. 32, Part 7-4-3 W6, Valladolid, Spain, 3-4
June, 1999.
Horning, N., 2010. Random Forests: An Algorithm for Image
Classification and Generation of Continuous Fields Data Sets.
Proceeding of International Conference on Geoinformatics for
Spatial Infrastructure Development in Earth and Allied
Sciences .
Jensen, J.R., 2005. Introductory Digital Image Processing, 3rd
Ed. Prentice Hall, Upper Saddle River, N.
Joelsson, R., Benediktsson, A. and Sveinsson, R., 2008.
Random Forest Classification of Remote Sensing Data. In: C.
Chen, ed. Image Processing For Remote Sensing. Boca Racon,
FL: CRC Press, pp. 61-78.
1962. IEEE
pattern | discrimination,
Julesz, B. Visual
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B7, 2012
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia
Transaction on Information Theory, 8:84-92.
Perona P., and Malik, J., 1990. Scale-space and edge detection
using anisotropic diffusion, IEEE Transactions on Pattern
Analysis and Machine Intelligence, 12 (7): 629—639.
Stumpf, A. and Kerle, N., 2011. Object-oriented mapping of
landsides using Random Forests, Remote Sensing of
Environment, 115:2564-2577.
Wang, X., Waske, B. and Benediktsson, A., 2009. Ensemble
Methods for Spectral-spatial Classification of Urban
Hyperspectral Data. In Proceedings of IGARSS, Volume 4, pp.
944-947.
ACKNOWLEDGEMENTS
This study was financially supported by a NSERC discovery
grant that was awarded to Prof. Dr. Jonathan Li at the
University of Waterloo.
39 Features
48 Features
Kappa = 0.551 Kappa =0.565
36 Features | 30 Features
Kappa =0.573 _ Kappa =0.689
26 : Features 22
Abe
b | a
a =0.681
Kappa =0.756
Kapp
15 Features
Kappa =(.726
Figure 5. the Maximum Likelihood classification results based on feature selection of Random Forests