PSCLOPOT

Description

Parameters

Parameter descriptions

FILI

The name of the input polarimetric SAR data set, which must be either non-symmetrized or symmetrized fully polarimetric (quad-polarization) complex data. The input data set must be in covariance, coherence, or Kennaugh matrix format.

FILI must be a data set that has already been imported into the PCIDSK (.pix) format by SARINGEST.

The input data set should have an equivalent number of looks (ENL) of at least 25, which could be achieved by applying a boxcar filter (see PSBOXCAR) or adaptive Lee filter (see PSPOLFIL) when the input data set is single-look complex.

CBDT

Optionally specifies the name of the text file that provides the class boundary definitions. By default, the standard 16 Cloude-Pottier classes are used, but the number of classes and class boundaries can be modified, as long as the text file conforms to the specified format.

Each line in the text file contains the following values, separated by one or more space or tab characters; the Class_Color_*, Class_Name, and Class_Description entries are optional:

[Class_Number] [Entropy_Min] [Entropy_Max] [Alpha_Angle_Min] [Alpha_Angle_Max] [Anisotropy_Min] [Anisotropy_Max] [Class_Color_Red] [Class_Color_Green] [Class_Color_Blue] [Class_Name] [Class_Description]

1 0.9 1.0 55.0 90.0 0.5 1.0 244 26 62 "Zone 1" "High Entropy, Anisotropic, Multiple Scattering"
6 0.0 0.5 47.5 90.0 0.5 1.0 26 118 244 "Zone 2" "High Entropy, Anisotropic, Volume Scattering"
10 0.9 1.0 40.0 55.0 0.0 0.5 27 158 33 "Zone 3" "Medium Entropy, Anisotropic, Multiple Scattering"
16 0.0 0.5 0.0 42.5 0.0 0.5 255 210 0	"Zone 4" "Medium Entropy, Anisotropic, Volume Scattering"

If multiple classes overlap, the first defined class to meet the specifications is chosen. If no classes meet the specification, the pixel is assigned class "0" (unknown). The class number of any class must be a unique value between 1 and 255 (inclusive).

In the example above, there would be four defined output classes with values 1, 6, 10, 16 and one undefined class labeled "unknown".

1 0.9 1.0 55.0 90.0 0.5 1.0 40 60 0 "Zone 1" "High Entropy, Anisotropic, Multiple Scattering" 
2 0.9 1.0 40.0 55.0 0.5 1.0 0 88 22 "Zone 2" "High Entropy, Anisotropic, Volume Scattering" 
3 0.5 0.9 50.0 90.0 0.5 1.0 227 128 0 "Zone 3" "Medium Entropy, Anisotropic, Multiple Scattering" 
4 0.5 0.9 40.0 50.0 0.5 1.0 0 255 17 "Zone 4" "Medium Entropy, Anisotropic, Volume Scattering" 
5 0.5 0.9 0.0 40.0 0.5 1.0 0 255 255 "Zone 5" "Medium Entropy, Anisotropic, Surface Scattering" 
6 0.0 0.5 47.5 90.0 0.5 1.0 255 0 0 "Zone 6" "Low Entropy, Anisotropic, Multiple Scattering" 
7 0.0 0.5 42.5 47.5 0.5 1.0 255 255 0 "Zone 7" "Low Entropy, Anisotropic, Dipole Scattering" 
8 0.0 0.5 0.0 42.5 0.5 1.0 0 0 255 "Zone 8" "Low Entropy, Anisotropic, Surface Scattering" 
9 0.9 1.0 55.0 90.0 0.0 0.5 126 144 0 "Zone 9" "High Entropy, Isotropic, Multiple Scattering" 
10 0.9 1.0 40.0 55.0 0.0 0.5 0 171 43 "Zone 10" "High Entropy, Isotropic, Volume Scattering" 
11 0.5 0.9 50.0 90.0 0.0 0.5 255 212 84 "Zone 11" "Medium Entropy, Isotropic, Multiple Scattering" 
12 0.5 0.9 40.0 50.0 0.0 0.5 139 255 148 "Zone 12" "Medium Entropy, Isotropic, Volume Scattering" 
13 0.5 0.9 0.0 40.0 0.0 0.5 83 191 255 "Zone 13" "Medium Entropy, Isotropic, Surface Scattering" 
14 0.0 0.5 47.5 90.0 0.0 0.5 255 112 112 "Zone 14" "Low Entropy, Isotropic, Multiple Scattering" 
15 0.0 0.5 42.5 47.5 0.0 0.5 255 255 112 "Zone 15" "Low Entropy, Isotropic, Dipole Scattering" 
16 0.0 0.5 0.0 42.5 0.0 0.5 138 168 255 "Zone 16" "Low Entropy, Isotropic, Surface Scattering"

Because of the way the alpha-entropy space is defined in the default Cloude-Pottier boundary definitions, pixels with entropy values between 0.9 and 1.0 and alpha angles between 0 and 40 are assigned to class 0 (unknown).

It is important to keep in mind that although there is some degree of unpredictability in how the boundaries between these classes are defined, they are closely linked with physical scattering mechanisms, independent of the training data set.

FILO

Specifies the name of the output classified image file. The output file has the same dimensions as the input file, and one 8-bit channel, where its pixel values specify the class number that is assigned to each pixel. Zero-valued output pixels represent input image pixels where the computation of the entropy, alpha angle, and anisotropy parameters falls into an undefined class range as defined by the class definitions or default Cloude-Pottier classification.

The specified file must not already exist.

MONITOR

The program progress can be monitored by printing the percentage of processing completed. A system parameter, MONITOR, controls this activity.

Available options are:

• ON: Turns the monitor ON (default)
• OFF: Turns the monitor OFF (recommended if running in batch or background mode)

Details

PSCLOPOT performs an unsupervised Cloude-Pottier image classification of a POLSAR data set. The classification can be applied to filtered, non-symmetrized or symmetrized fully polarimetric (quad-polarization) data. The classification is based on the partitioning of the entropy, alpha angle, and anisotropy feature space described in the article cited in the 'References' section. The extraction of the entropy, alpha, and anisotropy parameters is described in the Help for the PSEABA function.

The input data set must contain either non-symmetrized or symmetrized, filtered, fully polarimetric (quad-polarization) data in one of the following matrix formats: covariance (c4r6c or C3r3c), coherency (t4r6c or T3r3c), or Kennaugh (k16r or K9r).

The georeferencing segment and other segment types are copied from the input file to the output file.

Example

In the following example, perform a classification of an SLC RADARSAT-2 image already ingested by SARINGEST. The SLC image is first averaged by the PSBOXCAR program using a 5x5 window. PSBOXCAR also converts the image from the original non-symmetrized scattering matrix (s4c) format to the non-symmetrized covariance matrix format (c4r6c). The averaged image is then used as an input to PSCLOPOT with customized class boundaries.

EASI>FILI="rsat2_slc.pix" 
EASI>FILO="r2_enl25.pix"
EASI>FLSZ=5,5

EASI>run PSBOXCAR

EASI>FILI="r2_enl25.pix"
EASI>CBDT="cp_classes.txt"
EASI>FILO="r2_clopot.pix"

EASI>run PSCLOPOT

Algorithm

For every pixel, PSCLOPOT compares the computed values of the entropy (H), alpha (a), and anisotropy (A) with the class boundaries. The pixel is assigned to the first class for which the three inequalities are simultaneously satisfied:

Hmin <= H < Hmax
amin <= alpha < amax
Amin <= A < Amax

Pixels that do not fall into a defined class are assigned a value of "0 (unknown)".

The processing in PSCLOPOT is similar to that in PSEABA. The two functions differ in that PSCLOPOT performs an unsupervised classification in the entropy, alpha angle, and anisotropy space and creates a classification map, while PSEABA provides access to the individual parameter values at every pixel.

Acknowledgements

PCI Geomatics gratefully acknowledges the financial support provided by the Canadian Space Agency through the Earth Observation Application Development Program (EOADP), contract number 9F028-034946.

References

Cloude, S.R., and E. Pottier. "An entropy based classification scheme for land applications of polarimetric SAR", IEEE Trans. Geosci. Remote Sensing, 35, no. 1 (1997): 68-78.