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.