ANSI-C program: radcal_pwr_stat.c
NAME
rad_pwr_stat - radiometrically calibrate a stack of SLC images
using points with low temporal variablility.
SYNOPSIS
radcal_pwr_stat <SLC_tab> <SLC_par>
<plist> [MSR_cal] [PWR_cal] [roff] [loff] [nr] [nl]
[plist_out]
| <SLC_tab> | (input) two column list of the SLC filenames and SLC parameter filenames of the uncalibrated SLC images |
| <SLC_tab_cal> | (input) two column list of the SLC filenames and SLC parameter filenames of the calibrated SLC images (enter - for none) |
| <plist> | (input) point list for the point to use for calibraton (int, enter - to use the data to determine the calibration points) |
| [MSR_cal] | mean/sigma ratio for point target selection for relative calibration between scenes: 1.500 |
| [PWR_cal] | intensity threshold ratio for point target selection for relative calibration between scenes: 1.000 |
| [roff] | offset to starting range of section to analyze (default -: 0) |
| [loff] | offset to starting line of section to analyze (default -: 0) |
| [nr] | number of range pixels to analyze (default -: to end of line) |
| [nl] | number of azimuth lines to analyze (default -: to end of file) |
| [plist_out] |
point list of points used to determine calibration
using MSR_cal and PWR_cal thresholds |
EXAMPLE
radcal_pwr_stat SLC_tab SLC_tab_cal - 1.5 1.0 - - - -
plist_cal
Performs relative calibration of the SLCs listed in SLC_tab
and writes the calibrated output to the SLCs listed in
SLC_tab_cal. The data are used to determine points with low
temporal variablity. These points are then used to
determine relative calibration scale factors for the SLC images.
The coordinates of these points are written to the output point
list plist_cal.
DESCRIPTION
radcal_pwr_stat detects points with low temporal variability in
the image stack and uses these points to perform relative
radiometric calibration. The assumption is that points with low
temporal variablity are radiometrically stable and can be used
for relative calibration. The temporal variability is measured by
the ratio of the temporal mean to sigma ratio, where sigma is the
standard deviation of the backscatter intensity.
Alternately the user can provide a list of point coordinates using the plist parameter. An important condition for a successful use of radcal_pwr_stat is accurate co-registration of the SLCs.
The sum of all the point intensities in the list is evaluated
for each SLC at the point coordinates. The ratio of the SLC sum
for the points to the average power at the points of all
the layers is evaluated for each SLC to calculate a radiometric
scale factor. The square root of this scale factor is then
applied to the SLC real and imaginary components and the output
is written to the SLC listed in the SLC_tab_cal.
There must be the same number of entries in the SLC_tab_cal as in
the SLC_tab.
The physical basis for the methodology used is that point targets (i.e. SAR image resolution cells for which the scattering is dominated by a target which is small in size as compared to the size of the resolution cell) do not exhibit the speckle observed for extended targets. For a point target almost the same backscattering intensity is found when observing from slightly different directions. For an extended target the backscattering will strongly fluctuate if the change in direction corresponds to an offset of the order of the critical baseline known in interferometry. For point targets which remain unchanged over time it is expected that the temporal variability is small relative to the temporal average. For extended targets with a fully developped speckle behavior it is expected that the temporal variability is of the order of the temporal average.
radcal_pwr_stat uses the Mean to Standard deviation Ratio
(MSR) as the (inverse) measure of the temporal variability. For
fully developed speckle the MSR should be 1.0. Lower temporal
variability corresponds to higher MSR values, so that 1.5 is good
place to start. As a second criteria for calibration point
selection is given by the PWR_cal threshold that sets an minimum
SLC intensity to accept a point as a candidate. For very
small sets of SLC the intensity criteria may still be useful
while the MSR criteria may get more unreliable due to poor
statistics. To get reliable statistics with temporal variability
a sufficient number of scenes is required with a minimum of about
20.
Instead of the full area of the SLC a sub-section can be selected using the roff, loff, nr, and nl command line parameters.
It is also possible to use this program iteratively and it
should cause the relative calibration factors to converge. To
iterate the relative calibration use the calibrated output SLCs
as the input data stack. Then determine a new set of points with
low temporal variablity. There should be more points than
initially were found because the scenes are hopefully better
radiometrically calibrated. The calibration factors determined by
radcal_pwr_stat should should rapidly converge to 1.
The calibration factor stored in the SLC parameter file is
updated for the applied relative scaling.
The file names of the co-registered SLC are provided in a text file (ascii) with each line containing one SLC file name in the first column and the SLC parameter file name in the second column. A possibility to generate this text file is to use a command mk_tab.
SEE ALSO
mk_tab