NAME
r.affinity - An raster function that classifies remote
sensing images as well as other kinds of maps.
(GRASS Raster Processing Program)
SYNOPSIS
r.affinity
r.affinity help
DESCRIPTION
r.affinity is an analysis classifier based on affinity
index of David W. Goodall, which is quite different with
other indices by taking into account the order of the value
instead of the value itself. See AFFINITY BETWEEN AN
INDIVIDUAL AND A CLUSTER IN NUMERICAL TAXONOMY by David W.
Goodall 1968 for details. The advantage of this program is
that it can deal with multi source data. Three kinds of data
are supported by this program:
1. quantitative map such as image or DTM data
2. qualitative data such as a landcover map
3. ranked data such as a slope class map
It should be used as a supervised classification which
determines to which category each cell in the image
has the highest probability of belonging based on training areas
(groups of image pixels) chosen by the user.
The raster map layer output by r.affinity is a classified
image in which each cell has been assigned to a spectral
class (i.e., a category). The spectral classes (categories)
can be related to specific land cover types on the ground.
The program will run only interactively.
The user should first create the rast map with all the training areas
using i.digit or in other way. Then simply type r.affinity in the
command line without program arguments.
Parameters:
- training_map The rast map with all the training areas
- group The imagery group contains the subgroup to
be classified.
- subgroup The subgroup contains image files and maps, which
are goning to be classified.
- class_map The name of a raster map holds the
classification results. This new raster
map layer will contain categories that can
be related to land cover categories on the
ground.
- reject_name The optional name of a raster map holds
the reject threshold results. It is the
reject threshold map layer, and contains
one calculated confidence level for each
classified cell in the classified image.
Generally those cells whose confidence
level is less than 0.05 can be reguarded
as well classified. One of the possible
uses for this map layer is as a mask, to
identify cells in the classified image
that have the lowest probability of being
assigned to the correct class.
- data_type The data type for every file within subgroup
input one number for each file.
1. quantitative data
2. quanlitative data
3. ranked data
NOTES
The affinity classifier assumes that the attributes used in
classification are independent with each other. It is better
to choose those images and maps which are related as less as
possible.
This program respects the region and the mask set by the user.
SEE ALSO
Grass Tutorial: Raster Processing
r.digit i.group, r.mask
AUTHOR
Tian Qing, Institute of Remote Sensing Application, Chinese
Academy of Science
ACKNOWLEDGEMENTS
I wrote this program in Department of Biology, University of
Trieste, Italy under a fellowship from ICS/UNIDO .
My superviser is Prof. Enrico Feoli (Department of Biology,
University of Trieste, Italy)
I also would like to thank Prof. Paola Ganis (Department of
Biology, University of Trieste, Italy) and Dr. Cristina
Milesi (Centre for Theoretical and Applied Ecology) etc.
NOTICE
This program is part of the contrib section of the GRASS
distribution. As such, it is externally contributed code
that has not been examined or tested by the Office of GRASS
Integration.