Mapping soil types from multispectral scanner data.Multispectral remote sensing and computer-implemented pattern recognition techniques were used for automatic ?mapping' of soil types. This approach involves subjective selection of a set of reference samples from a gray-level display of spectral variations which was generated by a computer. Each resolution element is then classified using a maximum likelihood ratio. Output is a computer printout on which the researcher assigns a different symbol to each class. Four soil test areas in Indiana were experimentally examined using this approach, and partially successful results were obtained.
Document ID
19720028215
Acquisition Source
Legacy CDMS
Document Type
Conference Proceedings
Authors
Kristof, S. J.
Zachary, A. L. (Purdue University Lafayette, Ind., United States)
Date Acquired
August 6, 2013
Publication Date
January 1, 1971
Subject Category
Geophysics
Meeting Information
Meeting: International Symposium on Remote Sensing of Environment, 7th, University of Michigan