Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4465198 | International Journal of Applied Earth Observation and Geoinformation | 2010 | 4 Pages |
Abstract
This letter shows how conventional methods for satellite image classification can be improved by applying some filtering algorithms as a pre-classifying step. We will introduce a filtering scheme based on convolution equations of fractional type. The use of this kind of filter as a pre-classification step will be illustrated by classifying MODerate-resolution Imaging Spectroradiometer (MODIS) data to map burned areas in Mediterranean countries. The methodology we propose improved the estimations obtained by merely classifying the post-fire images (i.e. without filtering) in the study areas considered.
Keywords
Related Topics
Physical Sciences and Engineering
Earth and Planetary Sciences
Computers in Earth Sciences
Authors
C. Quintano, E. Cuesta,