Grand Challenge for Land Cover Dynamics
Classification maps generated by mixture modeling
This slide illustrates the application of a new algorithm for solving
the mixture modeling problem to a remotely sensed image of part of
Africa. Mixture modeling allows environmental scientists to estimate
the proportions of different vegetation types present in a single
pixel, thereby characterizing the vegetation more realistically than a
classification result that labels each pixel as a single vegetation
type. Accurate descriptions of the land surface are important boundary
conditions for climate models and other types of global environmental
models. The mixture modeling problem involves estimating the
proportions of different vegetation types from remotely sensed images.
These proportions are estimated by comparing the observed reflectance
measurements within a pixel to the expected measurements one would
obtain if the pixel were purely of one ground type, and solving for
the proportions using mathematical optimization procedures. The
classical methods employed by environmental scientists to solve this
problem suffer from a variety of numerical instabilities and
computational deficiencies; while the new algorithm developed at
Maryland - based on solutions to similar problems in image restoration
- provide results that are more accurate using algorithms that are faster.
Next Page
This page maintained by:ctso@umiacs.umd.edu