ISIS 3 Application Documentation
Create a digital elevation model from one or more control network files.
Description
The cnet2dem applicaton produces an interpolated digital elevation model (DEM) from
one or more control networks. There are fourteen different algorithms that
may be selected that provide either absolute radius values or other information,
such as radius and point distance statistics, computed at each output DEM
pixel value. The range and type of the DEM generated from cnet2dem is
a standard ISIS map definition or a projected ISIS cube file containing a
map definition. You can select any number of algorithms in the ALGORITHM
parameter. The output cube will contain the same number of bands in the
same order algorithms are specified in the parameter. The BandBin group
Name parameter will also contain the name of the algorithm that
cooresponds to the data contained in each band.
Note that the control network is expected to be the output results of
the bundle adjustment computed by jigsaw. Additionally,
jigsaw must have been used to solve for radius. Otherwise, you
will be using the apriori radius which is typically the PCK defininition
(will will result in the ellipsoid).
The adjusted latitude/longitude coordinate is used to compute the
location of all the control points. The adjusted coordinates are
compute by jigsaw. If it doesn't exist, the apriori coordinate
will be used but know that this coordinate is typically computed from
each reference image in each control point (so it may not be
consistent).
Also, in some cases, one or more of the algorithm radius bands in the
output of cnet2dem may be sufficient for direct use as an ISIS
shape model (after demprep is run!). However, this is generally
not the case and some of the radius bands may require a series of
filters to smooth out the results.
Categories
History
Kris Becker  20151113 
Original Version

Tyler Wilson  20160310 
Minor documentation corrections.


Parameter Groups
Input
Name

Description

CNET  Master control network file to match. 
CNETLIST  List of merger control network files. 
MAP  Specifies a file containing the output map specfications. 
Output
Name

Description

TO  Output DEM ISIS cube 
TRIM 
Null all pixels outside lat/lon boundaries.

Parameters
Name

Description

SEARCH  Optional tolerance for radial search range (meters). 
DISTANCE 
Specifies the radial search distance to search.

KDNODES  Specify the leaf size of the kdtree. 
NEIGHBORS  Specify the number of nearest NEIGHBOR points 
MINPOINTS  Specify the minimum number of search points per pixel. 
SIGMARADIUS 
Specifies noise tolerance in radius.

ALGORITHM  Type of data to provide a each output pixel location. 

Output:
TO
Description
Specifies the name of an output ISIS cube file that is created to
write the DEM. The output DEM may require demprep to be run
on it before it can be used as an ISIS DEM shape model.
Type
 cube 
File Mode
 output 
Pixel Type
 real 
Filter
 *.cub 
Output:
TRIM
Description
If this option is selected, pixels outside the latitude/longtiude
range will be trimmed (set to null).
This is useful for certain projections whose lines of latitude and
longitude are not parallel to image lines and sample columns.
Type
 boolean 
Default
 FALSE 
Parameters:
DISTANCE
Description
Radial distance in meters to search for control points measured
from the center of the ouput map pixel.
Type
 double 
Internal Default
 None 
Parameters:
KDNODES
Description
Number of leafs in the kdtree structure.
Parameters:
NEIGHBORS
Description
Specify number of control points that will be selected in the
nearest NEIGHBOR search query to produce the output DEM pixel
values.
Parameters:
MINPOINTS
Description
Specify number of minimum number of points to accept from the RADIAL
search query of the point cloud at each output DEM pixel. If the
result set for each pixel does not have at least MINPOINTS, NULL
values will be written to the output DEM pixel.
Parameters:
ALGORITHM
Description
This parameter provides numerous methods to compute the
output radius value for a given latitude, longitude and map grid
point. Any number of methods can be specified by enetering
a commadelimited lisf of the method names.
DEM Data Value Output Function Options
Name 
Parameters 
Description 
AverageRadius 
None 
For all control points found in the RADIAL and/or
NEARESTNEIGHBOR kdtree search, computes the average radius
value (the average Euclidean distance of all points from the surface
to the center of the body).

MaximumRadius 
None 
Computes the maximum radius (in meters) of the control points.

MinimumRadius 
None 
Computes the minimum radius (in meters) of all control points.

MedianRadius 
None 
Computes the median radius (in meters) of all control points.

StandardDeviationRadius 
None 
Computes the standard deviation (in meters) of the radii of
the control points.

MaximumDistance 
None 
Computes the maximum surface distance (in meters) of all control points
with respect to the latitude, longitude coordinates of the central pixel.

MinimumDistance 
None 
Computes the minimum surface distance (in meters) of all control points
with respect to the latitude, longitude coordinates of the central pixel.

MedianDistance 
None 
Computes the median surface distance (in meters) of all control points
with respect to the latitude, longitude coordinates of the central pixel.

AverageDistance 
None 
Computes the average surface distance (in meters) of all control points
with respect to the latitude, longitude coordinates of the central pixel.

StandardDeviationDistance 
None 
Computes the standard deviation (in meters) of the surface distance of
all control points with respect to the latitude, longitude coordinates
of the central pixel.

FrankeNelsonRadius 
None 
Computes the radius of the control points (in meters)
using the FrankeNelson method.

ShepardsRadius 
None 
Computes the radius of the control points (in meters)
using Shepard's weighted algorithm.

NaturalNeighborRadius 
None 
Computes the radius of the control points (in meters)
using the Nearest Neighbor interpolation algorithm.

TypeCount 
None 
Returns the total number of pixels in the point collection. It
additionally applies a signed (+/1) value to indicate
if the FLANN (Fast Library for Approximate Nearest Neighbors) search type
was nearest neighbor (1) or radius (+1).
