GAMMA Interferometric Point Target Analysis Software (IPTA): Reference Manual

sp_stat

ANSI-C program: sp_stat.c

NAME
sp_stat - generate point list using low spectral diversity to identify point targets

SYNOPSIS
sp_stat <SLC> <pwr> <cc> <MSR> <plist> <width> [PWR_min] [CC_min] [MSR_min] [rlks] [azlks] [roff] [loff] [nr] [nl] [bx] [by] [type] [r_ovr]

EXAMPLE
sp_stat 08456.rslc 08456.sp_stat.pwr 08456.sp_stat.cc 08456.sp_stat.MSR testsite.plist 1000 10. 0.4 1.2 4 4 - - - - 4 4 1

Generates a point list of coordinates of points with a low spectral diversity (i.e. high spectral correlation value) and a dominant intensity. The image files 08456.sp_stat.pwr, 08456.sp_stat.cc, and 08456.sp_stat.MSR are generated for visual control of the features used in the identification of the points.

INSTALLATION
Source code sp_stat.c in ./IPTA/src, executable version sp_stat in ./IPTA/bin

DESCRIPTION
In the IPTA point lists contain the coordinates (range and azimuth pixel number relative to the reference SLC geometry) of the selected points. Data at these points (SLC values, interferogram values, ...) are stored in point data stacks with each record corresponding to a different SLC or interferogram. The point data stack of mask values contains for each point a flag (0 or 1) which indicates if the data of a point are considered (1) or not considered (0).

sp_stat supports the generation of lists of points with "point target characteristics". This is done based on a single SLC image based on the spectral diversity. Only for point targets it is expected that the energy remains more or less the same when processing different looks with fractional azimuth and range bandwidth. Based on this low spectral diversity, respectivity based on correlation values above the indicated threshold potential point targets can be identified. As additional criteria a dominent intensity, i.e. pwr relative to an average power is above a given threshold, and a high mean to sigma ratio (MSR) between the spectral looks, can be used.

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 processing different looks with fractional azimuth and range bandwidth.

sp_stat uses the Mean to Standard deviation Ratio (MSR) as the (inverse) measure of the spectral variability. Lower spectral variability corresponds to higher MSR values. 1.2 might be a possible threshold to use.

As an additional (or alternative) criteria an SLC intensity above a minimum value can be indicated. As a simplification not an absolute value, but one relative to the spatial average is used. The idea behind this second criteria is to limit the points to dominating points (which need to be above the background level).

An alternative measure for the spectral diversity is the spectral correlation which is calculated in sp_stat. This measure is high for a low diversity. Concsequently, values above a given threshold are point target candidates. One advantage of the correlation measure is that it is "well focussed", meanning that in the case of a point target a high values is typically only found for one pixel, while the other criteria tend to be met also by some adjecent pixels.

In sp_stat the point list is determined from a single SLC. This means that no temporal criteria are considered. So the criteria used is clearly not temporal stability, but the point target characteristics expected in a single image.

To combine the functionality of sp_stat with multi-temporal criteria point lists sp_stat can be used to generated point lists for multiple co-registered SLC and then to combine these into a single list using the program merge_pt.

Advantages of sp_stat are that it can also be used in the case of a small serie of SLC (or even with a single SLC). Furthermore, it is less critical than pwr_stat, what concerns the radiometric calibration. On the other hand the SLC used should be well focussed.

A clear disadvantage of sp_stat is that the temporal stability criteria is not used. For an interferometric analysis of point targets it is essential that the point targets remain visible, which is only the case for a fraction of the point targets. A car, for example, may be identified as a point target, but it may not be present at the same location in a second SLC.

For large series of co-registered SLC pwr_stat is the preferred program to point lists. For small series sp_stat combined with a temporal consistancy element coming from the merging of several point lists using merge_pt is preferred.

Instead of the full area of the SLC a sub-section can be selected.

If the SLC has been oversampled in range, then the integer range over-sampling factor must be specified on the command line using the r_ovr parameter.

Other programs to generate new point lists are image2pt, mkgrid, merge_pt, prox_pt, and sp_stat.

Programs to validate existing point lists are qc_pt and expand_pt.

SEE ALSO
expand_pt, image2pt, merge_pt, mkgrid, prox_pt, pwr_stat, qc_pt, SLC_par, typedef_ISP.h, ipta.h.

© Copyrights for Documentation, Users Guide and Reference Manual by Gamma Remote Sensing, 2006.
UW, CW, TS, last change 1-May-2006.