Test 11: ReferencePeaks

Take PIT020 as the only maplet in MAPLIST.TXT. Generate 5cm images of the topography (which is based at TAG Site 1). Then we build a MAPLET, saving the MAPLETs flatMapVect at several states.

Simple Variations

Groups of images - (adjusted for 5cm)

• Test11A - Detailed Survey. Eric (Has 11 images, very low in sample space)
• Test11B - Min Orbit B (by hand, calculate the minimum requirement)
• Test11C - Average Orbit B (use a random Orbit B solution with 3 unique observation/illumination conditions)
• Test11D - Best Orbit B (as we did before)
• Test11E - Optimum Collection (as we did before)

• Test11U - Criss Cross (combine both of these sets)

Variations

Spacecraft Azimuth Variations

Sun Variations

Spacecraft Zenith Variations

Comparison

Hill Data

I have created the following hills for testing

• PIT007 - 7 deg tilt
• PIT014 - 14 deg tilt

• PIT020 - 20 deg tilt <-- Let's use this one

• PIT030 - 30 deg tilt
• PIT045 - 45 deg tilt
• PIT060 - 60 deg tilt

Truth Maplets

• -8.027 262.768
• T11TRU.MAP - already built. It has the truth heights loaded from PIT020.
• T11TRU.TXT -- Text file of the maplet. Use for plotting transits
• t11-truth.obj -- OBJ file for testing using CompareOBJ

Testing Maplets

• -8.027 262.768
• Build a model MAPLET using TAG Site 1 center position (5cm resolution and Q=49). Do not load a base map for topography
• Create MOD20#, which # is the test number (letter)
• Align the images with the key location at 50/50.
• Build template with 100
• (Skip until close to perfect) -- Do an auto align
• Build height using the iteration template
• Do flatMapVec and rename it to mod-20#-1
• Do compMapVec between MOD20# and T11TRU
• Repeat as necessary

Iteration Template

0
0
44
2
8                fill no data with shape slope
7
.01
6
y
y
1                condition with overlapping maps
2                condition with existing topo
.01
6
y
y
0                begin integration
.025
100
0
v
1
u
1
o                overlaps
RECENT
n
1
i
r
n
n

Useful copy/paste stuff

echo CENTER | /usr/local/localBin/flatMapVec 
 /usr/local/localBin/compMapVec MAPFILES/CENTER MAPFILES/T11TRU.MAP
~/eval/bin/Maplet2FITS MAPFILES/CENTER fit
 ~/eval/bin/FITS2OBJ --local fit center.obj
 ~/eval/bin/CompareOBJ    --compute-optimal-translation-and-rotation   center.obj newtru.obj 
convert  LMRK_DISPLAY1.pgm display.jpg
mkdir step02
cp CENTER.TXT MAPFILES/CENTER.MAP step02

gnuplot
set term x11
set style data lines
plot 'NEW020.TXT' u 0:50 lw 3 lc 0, 'CENTER.TXT' u 0:50 
set output 'out.png' 
set term png
replot
unset output
set term x11
exit

mv out.png mod20f.obj MOD20F.TXT comp.txt display.jpg step2

Save States

The goal is to see the evolution of the model over time. You can stop iterations once the peak and base are stable. The most likely tool to evaluate if the model has reached it best fit will be transit.

Save the flatMapVect statistics at the following number of iterations

• 1st
• 2nd
• 3rd
• 5th
• 10th
• 15th
• 20th
• 50th (if needed)
• 100th (if needed)
• 200th (if needed)

Examples

Example of a 14 degree peak, measured at difference locations

Example of a 20 degree peak, measured over time