== Autoregister == [[autoregister]] adds landmarks to one image at a time. Contrast that with [[lithos]] that adds images to landmarks. * Once you have a model, you will use [[autoregister]] to add images to the shape model. You will do this every time new images are collected. * [[autoregister]] does not update the topography itself. You still need to do an iteration to use the new images to build a template and generate topography. * When you accept the auto-alignment, SPC will update the spacecraft position and pointing. This is the same as running [[geometry]], option 2. * Look at the LMRK_DISPLAY1.pgm. How do the image/shape pair look. Any issues? If so, you may have to fix them. Autoregister the new MapCam images of the north pole. Do the following images (you registered these yesterday) {{{ M595949305F0 M595949731F0 M595950157F0 M595950583F0 M595951436F0 M595952715F0 M595953568F0 M595953994F0 M595956978F0 M595957405F0 M595959536F0 M595960815F0 }}} Basic autoregister cookbook: * autoregister * * y - add images * 0.5 * n * n - eliminate new * 0,50,.25,0,5 - criteria (notice there is one fewer parameter) * 1 - auto align * 2 - spacing * n - no new spacing * 0 - continue * y - accept changes (this will update the spacecraft pointing and position) * 1 - auto align * 1 - spacing * n - no new spacing * 0 - continue * y - accept changes (this will update the spacecraft pointing and position) * o - eliminate by correlation * .5 - Correlation value * 1 - auto align * 1 - spacing * n - no new spacing * 0 - continue * y - accept changes (this will update the spacecraft pointing and position and only use the remaining correlated images) * 4 - change flags * b - use all (this sets each image to be used when building a template) * 0 - end * q END == Tile a bigmap == Here you will tile the TAG site (Touch and Go site - where the spacecraft is planning on collecting a sample). We will do this in batch mode using make_scriptT. This uses a bigmap of the TAG site. The routine will put a new maplet every 50 pixels within the bigmap, and when it is done, it will have tile the entire area with a 30% overlap. Key differences between tiling using a bigmap and tiling using lat/lon * The position definition is based upon a reference bigmap * We use the topography of the bigmap Name this file script/make_scriptT_bigmap.in {{{ XXXXXX scripts/XXX075.SEED 50 150 100 150 150 150 50 100 100 100 150 100 50 50 100 50 150 50 END }}} Cookbook for this task: * Create the make_scriptT_bigmap.in in the scripts directory * Make a symbolic link {{{ ln -s scripts/make_scriptT_bigmap.in make_scriptT.in}}} * Copy the TILE00.MAP file into your MAPFILES directory. Normally, you'd create this yourself, but to save time and make things straight forward, I created it for you. It is almost exactly like the bigmap MTAG15 that you created yesterday, but the resolution has been selected to make correctly spaced maplets. [[attachment:TILE00.MAP]] * Make a symbolic link from TILE00.MAP to XXXXXX.MAP in the MAPFILES directory {{{ cd MAPFILES ln -s TILE00.MAP XXXXXX.MAP }}} * Double check all of your files. Running showmap on XXXXXX is a useful way to check that you have the right map installed. {{{ head make_scriptT.in head scripts/XXX075.SEED echo XXXXXX | showmap convert XXXXXX.pgm xxxxxx.jpg; open xxxxxx.jpg }}} * In the working directory, run make_scriptT * Review the output. You should see nine INN files. Check the file to see that it is correct * Run the run_script.b and monitor {{{ nohup sh run_script.b & find_nofitT }}} * When complete, rebuild the bigmap MTAG15 {{{ bigmap < lsupport/MTAG15.in echo MTAG15 | showmap convert MTAG15.pgm mtag15.jpg; open mtag15.jpg }}} * Check the create coverage of the newly created maplets using map_coverage. The first time, we are looking only at the 75m landmark. The second time, it is showing all landmarks (technically all landmarks between 0 and 150cm). {{{ map_coverage MTAG15 0 .00075 convert coverage_m.pgm 75.jpg; open 75.jpg map_coverage MTAG15 0 .00015 convert coverage_m.pgm 150.jpg; open 150.jpg } }}}