Environmentally-Oriented Processing of Multi-Spectral Satellite Images: New Challenges for Bayesian Methods
Remotely sensed images from new generation satellites present an opportunity for scientists to investigate problems in environmental and earth science which have been previously intractable. The magnitude of data that will arise from these hyperspectral instruments create the need for innovative techniques to accomplish data reduction. This paper presents an algorithm which shows promise as a tool for reducing the dimensionality of data resulting from remote sensing. The optimality criteria for the algorithm is the Bayes Risk in the reduced dimension space.
Keywordssatellite imaging multi-spectral satellite data environmental applications Bayes risk
Unable to display preview. Download preview PDF.
- 1.S.A. Starks, G.R. Keller, and D.E. Cooke, “Earth and Environmental Remote Sensing at PACES,” Geocarto, Int’l., 1997, Vol. 12, No. 3 (Sept. 1997).Google Scholar
- 2.T.M. Lillesand and R.W. Kiefer, Remote Sensing and Image Interpretation, J. Wiley, New York, 1994.Google Scholar
- 3.G. Asrar and R. Greenstone, eds., 1995 MTPE EOS Reference Handbook, NASA Goddard Space Flight Center, Greenbelt, MD, 1995.Google Scholar
- 4.Schott, Remote Sensing: The Image Chain Approach, Oxford University Press, New York, 1997.Google Scholar
- 5.R.J.R. De Figueiredo, “Optimal Linear and Nonlinear Feature Extraction from Several Gaussian Pattern Class,” Proc. of Second Joint Int’l Conf. On Pattern Recognition, Copenhagen, Denmark, August 1974.Google Scholar