CCDWISH

Description

CCDWISH measures change between two single-look complex SAR images based on modified Wishart statistics. The statistical probability of the area of interest being "the same" is computed within the defined analysis window. Areas with a high probability are considered to be unchanged, while areas with a low probability are considered to be changed; that is, they have a low probability of being the same.

Parameters

ccdwish(fili, filref, dbic, filo, winsize)

Name	Type	Caption	Length	Value range
FILI *	str	Input complex SAR image	1 -
FILREF *	str	Reference complex SAR image	1 -
DBIC	List[int]	Input complex SAR image channels	0 - 4
FILO *	str	Output Wishart change metric raster	1 -
WINSIZE	List[int]	Window size (pixels)	0 - 1	3 | 5 | 7 | 9 | 11 | 13 | 15 | 17 | 19 | 21 | 23 | 25 | 27 | 29 | 31 | 33 Default: 5

* Required parameter

Parameter descriptions

FILI

Specifies the name of the input file that contains the test SAR image data.

This data must be a complex SAR image in scattering matrix format (symmetrized or non-symmetrized) and it must have the same georeferencing, pixel size, and viewing geometry as the reference image file.

FILREF

Specifies the name of the file that contains the reference SAR image data.

The data must be a complex SAR image in scattering matrix format (symmetrized or non-symmetrized), and it must preserve the same structure as the input test image. This file must have the same georeferencing, pixel size, viewing geometry, bit depth, matrix representation, and calibration type as the specified input image.

DBIC

Optionally specifies the list of input channels to be processed. The specified channels must exist for both the input and reference images, and must be comparable; that is, they must have the same data type and polarization. By default, all channels from the input test and reference file are used to compute the change.

FILO

Specifies the name of the output file that is created. The output file will consist of the overlap area, and four channels that describe:

• the intensity value (or sum of intensities) of the input test data
• the intensity value (or sum of intensities) of the reference data
• the Wishart metric (written in decibels)
• the change, as percentiles, ranked from 0% (no change) to 100% (maximum change)

The direction of change is inferred by visual inspection of the output.

The specified file must not already exist.

WINSIZE

Optionally specifies the size of the square window that is moved over the images. All of the pixels in the window are averaged together for the value of the center pixel. The window size value must be an odd integer between 3 and 33. The default value of 5 defines a processing window of 5 pixels by 5 lines.

Details

CCDWISH uses Wishart statistics to test the equality of two complex covariance matrices that are assumed to follow a Wishart distribution. When the backscatter coefficients are very close, but either the phase or the correlation coefficient between the co-polarization have changed, the Wishart test statistic is much more sensitive to the differences than test statistics based only on the intensity. In other words, CCDWISH may highlight changes that are not detected by CCDINTEN.

Both the input and the reference images must have been ingested through SARINGEST and must overlap. It is recommended that you check the pixel alignment (or run AUTOSHIFT) before computing the change.

Example

This example uses two Radarsat-2 Fine QuadPol images created by SARINGEST to measure the changes between September 19th and September 25th, using Wishart statistics.

from pci.ccdwish import ccdwish

fili	=	'rsat2_FQ15_sep19.pix'	# input image file name
filref	=	'rsat2_FQ15_sep5.pix'	# reference image file name
dbic	=	[]		# use all channels to detect change
filo	=	'ccdwish_sep19_sep25.pix'	# output file name
winsize	=	[5]		# use a 5x5 a processing window

ccdwish(fili, filref, dbic, filo, winsize)
      

References

Conradsen, K.; Aasbjerg, A.; Schou, J.; Skriver, H.; "A Test Statisitc in the Complex Wishart Distribution and its Application to Change Detection in Polarimetric SAR Data" IEEE Trans. Geosci Remote Sensing, 41, no. 1 (2003): 4-19.