pkannogr(1)

NAME

pkannogr - classify vector dataset using Artificial Neural Network

SYNOPSIS

FCpkannogrF[0] FC-tF[0] training [FC-iF[0] input] [FC-oF[0] output] [FC-cvF[0] value] [options] [advanced options]

DESCRIPTION

pkannogr implements an artificial neural network (ANN) to solve a supervised classification problem. The implementation is based on the open source C++ library ( fann ⟨http://leenissen.dk/fann/wp/⟩ ). Both raster and vector files are supported as input. The output will contain the classification result, either in raster or vector format, corresponding to the format of the input. A training sample must be provided as an OGR vector dataset that contains the class labels and the features for each training point. The point locations are not considered in the training step. You can use the same training sample for classifying different images, provided the number of bands of the images are identical. Use the utility pkextract(1) to create a suitable training sample, based on a sample of points or polygons. For raster output maps you can attach a color table using the option FC-ctF[0].

OPTIONS

FC-iF[0] filename, FC--inputF[0] filename

input image

FC-tF[0] filename, FC--trainingF[0] filename

training vector file. A single vector file contains all training features (must be set as: B0, B1, B2,...) for all classes (class numbers identified by label option). Use multiple training files for bootstrap aggregation (alternative to the FC--bagF[0] and FC--bsizeF[0] options, where a random subset is taken from a single training file)

FC-tlnF[0] layer, FC--tlnF[0] layer

training layer name(s)

FC-labelF[0] attribute, FC--labelF[0] attribute

identifier for class label in training vector file. (default: label)

FC-priorF[0] value, FC--priorF[0] value

prior probabilities for each class (e.g., FC-priorF[0] 0.3 FC-priorF[0] 0.3 FC-priorF[0] 0.2 )

FC-cvF[0] value, FC--cvF[0] value

n-fold cross validation mode (default: 0)

FC-nnF[0] number, FC--nneuronF[0] number

number of neurons in hidden layers in neural network (multiple hidden layers are set by defining multiple number of neurons: FC-nnF[0] 15 FC-nnF[0] 1, default is one hidden layer with 5 neurons)

FC-mF[0] filename, FC--maskF[0] filename

Only classify within specified mask (vector or raster). For raster mask, set nodata values with the option FC--msknodataF[0].

FC-msknodataF[0] value, FC--msknodataF[0] value

mask value(s) not to consider for classification. Values will be taken over in classification image. Default is 0.

FC-nodataF[0] value, FC--nodataF[0] value

nodata value to put where image is masked as nodata (default: 0)

FC-oF[0] filename, FC--outputF[0] filename

output classification image

FC-otF[0] type, FC--otypeF[0] type

Data type for output image ({Byte / Int16 / UInt16 / UInt32 / Int32 / Float32 / Float64 / CInt16 / CInt32 / CFloat32 / CFloat64}). Empty string: inherit type from input image

FC-ofF[0] GDALformat, FC--oformatF[0] GDALformat

Output image format (see also gdal_translate(1)). Empty string: inherit from input image

FC-fF[0] OGRformat, FC--fF[0] OGRformat

Output ogr format for active training sample (default: SQLite)

FC-ctF[0] filename, FC--ctF[0] filename

colour table in ASCII format having 5 columns: id R G B ALFA (0: transparent, 255: solid)

FC-coF[0] NAME=VALUE, FC--coF[0] NAME=VALUE

Creation option for output file. Multiple options can be specified.

FC-cF[0] name, FC--classF[0] name

list of class names.

FC-rF[0] value, FC--reclassF[0] value

list of class values (use same order as in FC--classF[0] option).

FC-vF[0] 0|1|2, FC--verboseF[0] 0|1|2

set to: 0 (results only), 1 (confusion matrix), 2 (debug)

Advanced options

FC-balF[0] size, FC--balanceF[0] size

balance the input data to this number of samples for each class (default: 0)

FC-minF[0] number, FC--minF[0] number

if number of training pixels is less then min, do not take this class into account (0: consider all classes)

FC-bF[0] band, FC--bandF[0] band

band index (starting from 0, either use FC--bandF[0] option or use FC--startF[0] to FC--endF[0])

FC-sbandF[0] band, FC--startbandF[0] band

start band sequence number (default: 0)

FC-ebandF[0] band, FC--endbandF[0] band

end band sequence number

FC-offsetF[0] value, FC--offsetF[0] value

offset value for each spectral band input features: FCrefl[band]=(DN[band]-offset[band])/scale[band]F[0]

FC-scaleF[0] value, FC--scaleF[0] value

scale value for each spectral band input features: FCrefl=(DN[band]-offset[band])/scale[band]F[0]

FC-aF[0] 1|2, FC--aggregF[0] 1|2

how to combine aggregated classifiers, see also FC--rcF[0] option (1: sum rule, 2: max rule).

FC--connectionF[0] 0|1

connection rate (default: 1.0 for a fully connected network)

FC-wF[0] weights, FC--weightsF[0] weights

weights for neural network. Apply to fully connected network only, starting from first input neuron to last output neuron, including the bias neurons (last neuron in each but last layer)

FC-lF[0] rate, FC--learningF[0] rate

learning rate (default: 0.7)

FC--maxitF[0] number

number of maximum iterations (epoch) (default: 500)

FC-combF[0] rule, FC--combF[0] rule

how to combine bootstrap aggregation classifiers (0: sum rule, 1: product rule, 2: max rule). Also used to aggregate classes with FC--rcF[0] option. Default is sum rule (0)

FC-bagF[0] value, FC--bagF[0] value

Number of bootstrap aggregations (default is no bagging: 1)

FC-bsF[0] value, FC--bsizeF[0] value

Percentage of features used from available training features for each bootstrap aggregation (one size for all classes, or a different size for each class respectively. default: 100)

FC-cbF[0] filename, FC--classbagF[0] filename

output for each individual bootstrap aggregation (default is blank)

FC--probF[0] filename

probability image. Default is no probability image

FC-naF[0] number, FC--naF[0] number

number of active training points (default: 1)