Classify

Command line

The Classify operation can be directly executed by typing one of the following expressions on the command line of the Main Window:

  

OUTMAP

=

MapClassify(SampleSetName, ClassifierBox(factor))

 

OUTMAP

=

MapClassify(SampleSetName, ClassifierMinDist())

OUTMAP

=

MapClassify(SampleSetName, ClassifierMinDist(threshold))

 

OUTMAP

=

MapClassify(SampleSetName, ClassifierMinMahaDist())

OUTMAP

=

MapClassify(SampleSetName, ClassifierMinMahaDist(threshold))

 

OUTMAP

=

MapClassify(SampleSetName, ClassifierMaxLikelihood())

OUTMAP

=

MapClassify(SampleSetName, ClassifierMaxLikelihood(threshold))

 

OUTMAP

=

MapClassify(SampleSetName, ClassifierPriorProb(TableName, ColumnName))

OUTMAP

=

MapClassify(SampleSetName, ClassifierPriorProb(threshold, TableName, ColumnName))

where:

OUTMAP

is the name of your output raster map.

MapClassify

is the command to start the Classify operation.

SampleSetName

is the name of your input sample set which contains the training pixels; a sample set is created through sampling.

ClassifierBox

is obligatory syntax for a Box classification.

ClassifierMinDist

is obligatory syntax for a Minimum Distance classification.

ClassifierMinMahaDist

is obligatory syntax for a Minimum Mahalanobis Distance classification.

ClassifierMaxLikelihood

is obligatory syntax for a Maximum Likelihood classification.

ClassifierPriorProb

is obligatory syntax for a Maximum Likelihood classification which includes prior probabilities.

factor

is an obligatory parameter for the Box classifier which allows you to widen (factor > 1) the boxes that are 'drawn' around class means; the n-dimensional size of a box depends on the standard deviations in the n input bands; real value > 0.

threshold

for the Minimum Distance, Minimum Mahalanobis distance, Maximum Likelihood and Prior Probabilities classifier:

  • Optional parameter which is used, for a pixel to be classified, to decide whether the calculated spectral distance towards a class is small enough to actually classify the pixel as that class; if the spectral distance of a pixel towards a class is larger than the threshold, the pixel will not be classified as that class; real value > 0.
  • The larger you choose the threshold, the more pixels will be classified.
  • If the threshold is not specified, all pixels will be classified.

TableName

for the Prior Probabilities classifier: the parameter which specifies the table that contains the column with the prior probability values.

ColumnName

for the Prior Probabilities classifier: the parameter which specifies the column name that contains prior probability values.

When the definition symbol = is used, a dependent output map is created; when the assignment symbol := is used, the dependency link is immediately broken after the output map has been calculated.

See also: