dSARsim examples

Index

Urban area
Example 1, rotateBack
Example 2, changing the pixel spacing
Example 3, changing viewing direction
Mountainous area

Urban area

Let's take the following example, which represents some buildings in the northern part of the city of Trento, Italy. As required, the input file is in GeoTIFF format and in UTM coordinate system. Let's assume the input file is called urban_area.tif.

Some buildings in the northern part of the city of Trento, Italy.
Some buildings in the northern part of the city of Trento, Italy. Pixel spacing is 1 m.
Scale: blue = 186 m, red = 226 m.
These data are owned and freely distributed by Ufficio Sistemi Informativi, Servizio autorizzazioni e valutazioni ambientali, Provincia Autonoma di Trento. Downloaded from http://www.territorio.provincia.tn.it/.

Example 1, rotateBack

Issuing the command:
./dsarsim.py -i urban_area.tif -o example1.tif -ai 54 -aa 20 -s 4
we say that we want to pseudo-simulate the input file using an incidence angle of 54º, an aspect angle of 20º and 4 parallel processes. Implicitly, the viewing direction is from West to East (default). The result is the following:

Pseudo-simulation obtained using incidence angle = 54º and aspect angle = 20º. Viewing direction from West to East.
Pseudo-simulation obtained using incidence angle = 54º and aspect angle = 20º. Viewing direction from West to East.
Scale: black = 0, white = 4.

Don't forget that this is a slant range image. As the command didn't specify it, the output pixel spacing has been automatically calculated in order to obtain an output image comparable to the input one (i.e. in flat-ground range). That is, the slant range pixel spacing here is 0.81 m.
In this case you can use the -r option and obtain a "rotated back" image which is directly comparable with the input DSM. The output with the rotate back option is the following:

Pseudo-simulation obtained using incidence angle = 54º, aspect angle = 20º and the rotateBack option. Viewing direction from West to East.
Pseudo-simulation obtained using incidence angle = 54º, aspect angle = 20º and the rotateBack option. Viewing direction from West to East.
Scale: black = 0, white = 4.

Note: as specified in the README file, the rotateBack option should be used only when the output pixel spacing is calculated automatically (the user didn't set --azimuthPixelSpacing nor --slantRangePixelSpacing).

Example 2, changing the pixel spacing

In this example we want to simulate with the same viewing parameters but we set the output pixel spacing to be 1.2 m in the azimuth directions and 1.5 m in the slant range direction. The command to be issued is:
./dsarsim.py -i urban_area.tif -o example2.tif -ai 54 -aa 20 -pr 1.5 -pa 1.2
The output image is the following. You can note how, with respect to Example 1, it becomes smaller in both axis and sort of "compressed" in the slant range direction.

Pseudo-simulation obtained using incidence angle = 54º, aspect angle = 20º and azimuth and slant range pixel spacing equal to 1.2 m and 1.5 m, respectively. Viewing direction from West to East.
Pseudo-simulation obtained using incidence angle = 54º, aspect angle = 20º and azimuth and slant range pixel spacing equal to 1.2 m and 1.5 m, respectively. Viewing direction from West to East.
Scale: black = 0, white = 4.

Example 3, changing viewing direction

dSARsim can generate pseudo-simulations also from East to West. In this example we use the option -d e to tell dSARsim to change the viewing direction. We also set an incidence angle of 30º and an aspect angle of 12º. We use the default pixel spacing and the -r option so the image can be compared to that of Example 1. The command will thus be:
./dsarsim.py -i urban_area.tif -o example3.tif -ai 30 -aa 12 -d e -r
and the output is the following:

Pseudo-simulation obtained using incidence angle = 30º, aspect angle = 12º and the rotateBack option. Viewing direction from East to West.
Pseudo-simulation obtained using incidence angle = 30º, aspect angle = 12º and the rotateBack option. Viewing direction from East to West.
Scale: black = 0, white = 4.

Mountainous area

In this section examples similar to those showed for an urban area are reported for the case of a lower resolution DTM of a mountainous area. The input DTM represents the area South of the city of Trento, Italy:

DTM of the area South of the city of Trento, Italy.
DTM of the area South of the city of Trento, Italy. Pixel spacing is 30 m (after conversion from lat/long to UTM and resampling of the original data).
Scale: blue = 56 m, red = 2300 m.
These data are part of the ASTER GDEM and are freely distributed on http://gdem.ersdac.jspacesystems.or.jp/index.jsp. ASTER GDEM is a product of METI and NASA.

Incidence angle = 23º, aspect angle = 20º, viewing direction West to East

Pseudo-simulation obtained using incidence angle = 23º and aspect angle = 20º. Viewing direction from West to East.
Scale: black = 0, white = 4.

Incidence angle = 23º, aspect angle = 20º, viewing direction West to East and rotateBack option

Pseudo-simulation obtained using incidence angle = 23º, aspect angle = 20º and the rotateBack option. Viewing direction from West to East.
Scale: black = 0, white = 4.

Incidence angle = 23º, aspect angle = 20º, viewing direction West to East, azimuth and slant range pixel spacing = 35 m

Pseudo-simulation obtained using incidence angle = 23º, aspect angle = 20º and azimuth and slant range pixel spacing equal to 35 m. Viewing direction from West to East.
Scale: black = 0, white = 4.

Incidence angle = 21º, aspect angle = 12º, viewing direction East to West and rotateBack option

Pseudo-simulation obtained using incidence angle = 21º, aspect angle = 12º and the rotateBack option. Viewing direction from East to West.
Scale: black = 0, white = 4.