ANSI-C program: pt_density_reduction.c
NAME
pt_density_reduction - point density reduction to approx. target
point density
SYNOPSIS
pt_density_reduction <plist> <pmask>
<SLC_par> <pdata_in> <r_max> <pmask_out>
[tg_pt_dens] [rank_flag]
| <plist> | (input) point list (int) |
| <pmask> | (input) point data stack of mask values (uchar, set to - to accept all points) |
| <SLC_par> | (input) SLC parameter file of point list coordinates |
| <pdata_in> | (input) point data stack of quality values (type: float) |
| <r_max> | maximum radius (range samples, default: 64 ) |
| <pmask_out> | record number in pdata_in and pdata_out to process (default -: all records) |
| [tg_pt_dens] | target point density (number of points, default: 100) |
| [rank_flag] | flag defining the ranking order of input data for point reduction (0: decreasing (e.g. coherence), default=1: increasing (e.g. psigma)) |
EXAMPLE
pt_density_reduction testsite.plist testsite.pmask
08256.rslc.par psigma1 25 testsite.pmask_out 100 1
Determines a point mask such that the point density (= number of points in the indicated search radius, in the example 25 slant range pixels) is reduced to values approximately up to the indicated target point density (tg_pt_dens, 100 points in the example). The decision whether a valid input point (defined by the point list testsite.plist and the point mask testsite.pmask) is masked out is based on the target point density and the point quality value provided (psigma1). The quality value is used to preferably mask points of low quality.
INSTALLATION
Source code pt_density_reduction.c in ./IPTA/src, executable
version pt_density_reduction in ./IPTA/bin
DESCRIPTION
pt_density_reduction supports the reduction of the point target
density using a quality measure such as the phase standard
deviation psigma of the IPTA regression residuals (low psigma
value = high quality) or the coherence (high value = high
quality).
If a particular region of interest contains both built-up as well
as rural or vegetated area the point density distribution is very
likely to be spatially uneven: much more interferometric point
targets are found in the built-up area and along man-made
features.
Thus, the user might want to reduce the point density in such a
way that for areas with high point densities only the best points
are kept while the number of points for areas with point
densities that are lower than the target point density is kept.
Such an approach can be used to achieve a more optimal point
density distribution and to reduce the size of point lists
without much loss of information.
Together with the input point list an input point mask can be
indicated to consider or ignore specific points. "-" can be
indicated instead of a pmask file to consider all points of the
list.
The SLC parameter file associated with the point list coordinate
geometry is required to calculate ground distances between pairs
of points. Such distances are calculated using the ground range
sampling (calculated from slant range sampling and incidence
angle) and the azimuth spacing. The maximum search area radius is
indicated in SLC range samples. Only valid points, i.e. points
with a non-zero value in the mask are considered. For points with
a zero in the mask the output is set to the NULL value.
The program writes the search area size to the standard
output.
A single-record point data stack (data type: float) is required
as the quality measure based upon which the point density
reduction is performed.
The target point density (default: 100 points) indicates the
approximate max. point density that will be reached after point
density reduction. To get an idea on the point densities present
in the scene with a given search radius the program pt_density can be used.
The rank flag defines the quality measure based ranking order (0:
high value = high quality, default=1: low value = high quality).
The correct rank flag for the phase standard deviation is 1.
The result of pt_density_reduction is an output mask indicating
the reduced list of points.
The number of valid points after point density reduction is
written to the standard output.
To actually reduce the size of a point list file to a reduced
size can be done using the program cat_pt.
OPTIONS
pt_density_reduction can be used iteratively. A possible apporach
is to first reduce the point density using a small search window.
In a further step the output pmask of the previous step is used
as input pmask and the size of the search window is increased.
This permits avoiding too high point densities at different
spatial scales./p>
SEE ALSO
pt_density.c, cat_pt, ipta.h.
© Copyrights for Documentation, Users Guide and Reference Manual by Gamma Remote Sensing, 2011.
UW, CW, OF,
last change 29-Mar-2011.