Map projected CTX images ?, How to merge/overlay CTX with map projected HiRISE Options

[Nirgal]

Does anyone know if/where there are map projected CTX images avilable ?

There are the EDR (i.e. raw) data sets as IMGs at PDS, but so far there seem no RDRs (i.e. calibrated and map projected) versions of the CTX images publicly available (yet?)

However, when I'm looking at the tight CTX image integration in Google Mars it seems that they have the exact footprint data of all CTX images readily to be overlayed with the global Mars/MOLA/HRSC map ...

So somehow the map projection of CTX images/footprints must have been done at least for Google Mars.

Ross Beyer even has a link on his web page to the footprint data of all CTX images as KML files:
http://pirlwww.lpl.arizona.edu/~rbeyer/kml/CTX_PDS.kml

So I'm courious, if the complete map projected CTX image data exists somewhere or if it is required to do the projection for each image oneself with the usual ISIS tool chain only ?

Background: what I really want to do is to merge/overlay map projected HiRISE images with their corresponding CTX observation as a kind of "background"-layer ... Ideally this should be performed with pixel-precision which, of course, requires exact map projection from one
camera's space to the other.

Probably this must have been done before, but I couldn't find enough information on this in the ISIS doc & ISIS/GIS community forums.

Thanks in advance for any help !

Bernhard

[elakdawalla]

Bernhard, I imagine Doug will reply before long with tales of the pain he has gone through to reproject the CTX images, but it involves ISIS. The Google map overlays probably have nothing to do with the image files; I imagine they are simple polygons, rectangles made using the latitude/longitude coordinates of the four corners of each image. That lat/lon information is available in the headers of the images, of course.

If there are one or two images that might be of interest to a lot of people, I have a friend who I can beg to provide map-projected versions. But if you need more than that, you're going to have to figure out how to run ISIS, I think!

[djellison]

It's fairly easy - just follow this from end to end :
http://isis.astrogeology.usgs.gov/IsisWork...rbiter_CTX_Data

[Nirgal]

ok thank you Emily, Doug !

So it is definitely going to involve ISIS, as I have already suspected.

As far as the radiometric calibration is concerned I wrote my own C++ program some time ago using the source code in the ISIS program "ctxcal" as a reference -> that produced reasonable results converting the IMGs to calibrated 16 Bit Tiff/PNG.

However things are not so easy with map projection (without the SPICE kernels and such

So in the mean time I am in the process of downloading the whole ISIS installation (30+ Gigs ) to be run in a Linux VM....

The following quotes I found from the ISIS support forum might also be helpful:

CTX images are only released in raw PDS format. They have not been map projected which means they are not ready for a GIS or mapping application. All PDS raw (EDR) images must first be preocessed before they should be used. Some teams like HiRISE, MOLA, THEMIS, HRSC, CRISM, because their images are hard to work with or they are just being nice, release map projected images or mosaics as well as their PDS EDRs. These derived products, which again are map projected, can be run through pds2world.

The bottom line is that you will need to run these images through ISIS3 (or other app) to calibrate and push into a map projection.

Here is a CTX helper script that you can use once you have ISIS3 installed:
http://isis.astrogeology.usgs.gov/IsisSupp...opic.php?t=1393

I imagine Doug will reply before long with tales of the pain he has gone through to reproject the CTX images

It's fairly easy - just follow this from end to end :
http://isis.astrogeology.usgs.gov/IsisWork...rbiter_CTX_Data

@Doug: ok, so were you actually successful with projecting and stitching HiRISE and CTX together in the same image space ?

[tharrison]

CTX images are only released in raw PDS format. They have not been map projected which means they are not ready for a GIS or mapping application. All PDS raw (EDR) images must first be preocessed before they should be used. Some teams like HiRISE, MOLA, THEMIS, HRSC, CRISM, because their images are hard to work with or they are just being nice, release map projected images or mosaics as well as their PDS EDRs.

HiRISE, THEMIS, etc. provide map-projected images because they received funding from NASA to do so, not because they're being "nice." We (MSSS) did not receive any funding to release such products, or to produce a nice web interface from which to download them as we did with the MOC data. Unfortunately, that means that people have to slug through ISIS to use the data, and consequently the dataset is sorely underused by the planetary science community.

[djellison]

@Doug: ok, so were you actually successful with projecting and stitching HiRISE and CTX together in the same image space ?

Not done that - I'm not clever enough - I would simply use photoshop to lay HiRISE over CTX

[JohnVV]

the thread is a bit old but cartrite was doing some ctx images ( yes in isis3) over on celestia matters
http://forum.celestialmatters.org/viewtopic.php?t=194

the isis thread is a bit long 19 pages .

[Nirgal]

thanks for the hint, John !

the isis thread is a bit long 19 pages .

ok, but while the 30+ Gigabytes ISIS3 download is still running day & night (at only 100 KB/sec) I'm afraid I'll have enough time to read it

[Nirgal]

Meanwhile I found that I don't have to wait for the whole download to finish because all the files needed for CTX processing have already been transferred.

So I gave it a try and it seems to work so far

Here is my first ISIS-map projected CTX image, showing Oppys travel area between Victoria and Endeavour, re-projected with the same mapping parameters than the corresponding HiRISE-observation PSP_010341_1775.



(I possibly messed up the stretching of pixel brightness values somwhere during the processing chain, but the map projection itself seems ok so far)

Sequence of ISIS programs used is:

>mroctx2isis : transforms PDS IMG to ISIS Cube)
>spiceinit : prepare SPICE kernels
>maptemplate : create a map projection defnition file
>ctxcal : perform radiometric calibration
>ctxevenodd : destriping
>cam2map : map projection
>isis2std : transfrom ISIS cube to standard PNG format

Looks like exploring this ISIS universe is going to be a lot of fun