Celestia Forums

Posted: **16.09.2007, 06:59**

Hi all,
For those that would like to try to create their own textures, some time in the last 2 or 3 weeks a large amount of data was released through the Planetary Data System. There were about thirty more volumes released for the Context Imaging Camera on MRO and HiRise data is up to about 5000 or so orbits now. They also released a lot of data from ESA's MEX HRSC camera. Apparently this data is not only covering the visible spectrum but can also be used to create very detailed DEM's. So a lot of interesting images will probably become available soon in the near future.

I started to create mosaics from the CTX camera in 6 meter per pixel resolution to create a level 12 VT of the Vallesmarineris. The data is behind the Hirise releases and only covers up to orbit 2000 or so. There are still a lot of gaps in the coverage area but it is starting to fill in. I'm not sure if I'm going to try and combine data from MEX's HRSC camera or MRO's Crism camera and the HiRise / CTX cameras to produce true color textures before any release from me of a level 12 VT. With Fridger's tools out now I think if I do make any available, They will only be a base texture that will require further processing with F-tex tools. A preview of one of the mosaics I did of the Vallesmarineris can be seen at my site. Here are a couple of thumbs from tiles of the original resolution mosaic. It's coverage area is about 5 to 7 degrees south latitude and 288.5 to 290 degrees east longitude or around 70 to 71 west.

cartrite

Posted: **16.09.2007, 12:05**

pretty incredible...

Posted: **17.09.2007, 02:51**

They also released a lot of data from ESA's MEX HRSC camera. Apparently this data is not only covering the visible spectrum but can also be used to create very detailed DEM's.

I downloaded a group of files from http://hrscview.fu-berlin.de/cgi-bin/io ... =entry.ion
folder h2309 which had a file with dt4 extension. This was a dem for that image. Here are a couple of screenshots of that file after it was processed by ISIS3 to crop a smaller section and 3dem > usgs dem and wilbur.exe > .obj and then blender..

Here is a snapshot of the high resolution image level 3 that is easily opened with the gimp's pds plugin.

I also downloaded the level 4 high-res image and imported it to ISIS3.

credit HRSCview. Freie Universitaet Berlin and DLR Berlin, http://hrscview.fu-berlin.de/ which is where I found the above files.

cartrite

Posted: **18.09.2007, 02:36**

Pretty cool. BTW, in the two Blender screenshots there is an icon? object? that's to the left and below the cross hair icon - is that one of the Mars rovers?

Posted: **18.09.2007, 05:09**

That's my private spaceship that I fly around the galaxy with :wink:

but really it's the little bouncing icon, I like the onre that comes with the gimp. The brush in wilburs mouth is really comical. :lol:

Anyhow, a preview of things to come. I cropped out a part of the mosaic I did in my first post and matched it to a dem from MEX that was downloaded from Freie Universitaet Berlin and DLR Berlin. It's from folder h0334. Anyhow a few snapshots with my little spaceship in some if not all of the images. This is a down and dirty job cause I cropped out the texture with the free select tool. Still got to fine tune it.

cartrite

Posted: **01.01.2008, 16:14**

Here are a couple of screenshots of that file after it was processed by ISIS3 to crop a smaller section and 3dem > usgs dem and wilbur.exe > .obj and then blender..

I've grabbed a few of the .dt4.50's and HRSCView works great as an IDLVM package - the perspective views work well as well. However, if you export to VRML from the perspective view - the VRML files seem screwed. Okino Polytrans, Cosmo - nothing can load them as far as I can tell.

I'm desperate to find a way to get those .dt4's into 3ds Max in some way - any help at this point would be much appreciated!

Doug

Posted: **01.01.2008, 18:08**

djellison wrote:I'm desperate to find a way to get those .dt4's into 3ds Max in some way - any help at this point would be much appreciated!

After I decompress the dt4.50.bz2 to dt4.50 files, I use a couple of software packages to convert the dt4 files to a Digital Elevation Model (DEM). First I use 2 programs from the Integrated Software for Imagers and Spectometers (ISIS3) package, vicar2isis and isis2raw. If 3ds Max can load binaries then you should be good to go. You just need to rename the .raw file to .bin. You may need to open the dt4 file with a text editor to see what the number of lines, samples, and geo coordinates are.

If you need to load DEM's with 3ds MAX, the program 3Dem will convert the binaries to DEM's.

If you need wave objects .obj files to load into 3dsMAX, then the program Wilbur.exe will be able to load the DEM's and export as .obj files. It is here that I import the .obj files into Blender.

If you need 3ds files to load into 3ds MAX, Anim8or will import .obj files and export 3ds files. These dt4 files, unless they are cropped or under sampled, will produce files with well over 65000 triangles and Anim8or will not be able to export them.

I'm not familiar with 3ds MAX so I'm not sure what 3dsMAX can import and if it supports files with more than 65000 triangles in 3ds form, but I hope this helps.

This is
cartrite

Posted: **01.01.2008, 20:50**

I'm rsync'ing my way into ISIS3 as we speak - many thanks, I'll let you know how it goes!

Cheers

Doug

Posted: **01.01.2008, 21:00**

About ISIS, you don't need to download the whole data folder which is now about 30 GB and growing. For your purposes you may only need the base folder. The base folder is "only" about 2.5 GB and most of that is the 2 mola dems.
I think but this may have changed, This line will accomplish that.

Code: Select all

rsync -azv --delete isisdist.wr.usgs.gov::isis3data/data/base .

You would run that command from the same folder that you run the isis download command so when finished there will be an isis folder with the programs and a data folder side by side.
If you are running rsync then you must already have Linux or a MAC OS because they are the only platforms that ISIS will run. I Think.
GOOD LUCK
cartrite

Posted: **01.01.2008, 21:17**

That'll save me a days downloading probably

Thanks for that! The apps rsync had finished, I'd only just started the data rsync, that'll save be a load of downloading I wont need! Cheers.

Doug (on OSX)

Posted: **01.01.2008, 21:52**

Since you are using OSX, the problem now is that the other programs I provided links for only run with Windows. Hopefully 3dsMAX will import binary. If you look at the dt4.50.cub files by using the catlab program from isis, it will have the number of lines and samples that define a binary file for 3DEM which hopefully is similar to 3ds MAX.

cartrite

Posted: **01.01.2008, 22:44**

OSX for ISIS - a quad core XP64 box for everything else - hopefully I've got all the bases covered!

Thanks for all the help - I might be back for some more later in the week

Doug

Posted: **02.01.2008, 17:14**

It took a lot more rsyncing - just doing the mola stuff wasnt enough, there were some label templates and interface components that were hidden away.

However - I've succesfully vicar2isis and isis2raw'd - I'll be 3Dem'ing later!

Doug

Posted: **02.01.2008, 20:00**

Right - stuck at the 3dem point.

The header of the .dt4 is as follows..

LBLSIZE=3222 FORMAT='HALF' TYPE='IMAGE' BUFSIZ=65536 DIM=3 EOL=0 RECSIZE=3222 ORG='BSQ' NL=4802 NS=1611 NB=1 N1=1611 N2=4802 N3=1 N4=0 NBB=0 NLB=0 HOST='X86-LINUX' INTFMT='LOW' REALFMT='RIEEE' BHOST='VAX-VMS' BINTFMT='LOW' BREALFMT='VAX' BLTYPE='' PROPERTY='DTM' DTM_MISSING_DN=-32768 DTM_OFFSET=0.0 DTM_SCALING_FACTOR=1.0 DTM_A_AXIS_RADIUS=3396.0 DTM_B_AXIS_RADIUS=3396.0 DTM_C_AXIS_RADIUS=3396.0 DTM_DESC='HEIGHT_ABOVE_SPHEROID' DTM_MINIMUM_DN=-565 DTM_MAXIMUM_DN=2611 PROPERTY='FILE' PROCESSING_LEVEL_ID=4 FILE_NAME='h1201_0001.dt4.50' PROPERTY='MAP' TARGET_NAME='MARS' MAP_PROJECTION_TYPE='SINUSOIDAL' POSITIVE_LONGITUDE_DIRECTION='EAST' A_AXIS_RADIUS=3396.0 B_AXIS_RADIUS=3396.0 C_AXIS_RADIUS=3396.0 BODY_LONG_AXIS=0.0 MAP_SCALE=0.075 CENTER_LATITUDE=0.0 CENTER_LONGITUDE=25.0 SPHERICAL_AZIMUTH=0.0 CARTESIAN_AZIMUTH=0.0 LINE_PROJECTION_OFFSET=-3675.6 SAMPLE_PROJECTION_OFFSET=804.8 MAP_PROJECTION_DESC=('An equal-area, pseudocylindrical projection where the central meridian is a','straight line. All other meridians are shown as equally spaced sinusoidal curves.','Parallels are equally spaced straight lines, parallel to each other. Poles are','points. Scale is true along the central meridians and all parallels.','Equations (30-1),(30-2),(30-6),(30-7) of USGS Paper 1395 (pp 247,248)',' were used.','The value of the COORDINATE_SYSTEM_NAME item determines whether latitudes are','planetographic or planetocentric; if this keyword is absent, then','the default is planetographic.','The direction of increasing longitude is defined by the POSITIVE_LONGITUDE_DIRECTION','item; if this keyword is absent, then the direction is determined by','COORDINATE_SYSTEM_NAME: it is East if the system is planetographic, West if it is','planetocentric.','NOTE: Portions of above text taken from U.S. Geological Survey Professional Paper 1395,','second printing 1989, ''Map Projections - A Working Manual'' by John Snyder.','See text for detailed formulae.') COORDINATE_SYSTEM_NAME='PLANETOCENTRIC' TASK='DLRDTM' USER='gwinner' DAT_TIM='Thu Jul 12 09:05:46 2007' DATUMSHIFT_FILE='NOSHIFT' GPS_DOMAIN='MARS' DATUMSHIFT_FILE_HEIGHT='NOSHIFT' HREF_DOMAIN='self defined, see Property DTM' DTM_IPOL_MOD='NO_GAPS' TPC_KERNEL_FILE_NAME='CONSTANTS'

I've exported to 32 bit with isis2raw - so obviously FLOAT (32) for loading a terrain matrix in 3dem. LE or BE?

1611 columns, 4802 rows then - BUT - I'm screwed to fill in the rest of it - any ideas?

Once I get one thru the pipeline, I'll be sorted - it's just the first run that's the nightmare! Thanks for holding my hand so well - I've got further in getting HRSC into an animatable format in the last 24 hours than the last 6 months thanks to your help.

Cheers

Doug

Posted: **02.01.2008, 20:51**

This is the trouble I had. How to come up with the corner coordinates.
3DEM don't really care what those coordinates are as long as they are in the right order. If you input the wrong coordinates in the right order, the file will load but the exported dem will have the wrong coordinates. This is not that important as long as you fix them later. If you are going to use Wilbur to save them as waveobj files, you can fix them then. When I create the waveobj file, I enter the fractions for min xyz and max xyz which are between -1 and 1 for the area the dem is covering. This gets quite complicated.
But here are a couple of screen shots that show the use of 3dem for the file you were using.

cartrite

Posted: **02.01.2008, 21:17**

Ahhh - found some coords!

I cheated, went to the PDS HRSC library, where the map projected IMG's have great header info :

OBJECT = IMAGE_MAP_PROJECTION
^DATA_SET_MAP_PROJECTION_CATALOG= "DSMAP.CAT"
A_AXIS_RADIUS = 3396.19 <km>
B_AXIS_RADIUS = 3396.19 <km>
C_AXIS_RADIUS = 3396.19 <km>
CENTER_LATITUDE = 0.0
CENTER_LONGITUDE = 178.0
COORDINATE_SYSTEM_NAME = PLANETOGRAPHIC
COORDINATE_SYSTEM_TYPE = "BODY-FIXED ROTATING"
EASTERNMOST_LONGITUDE = 179.06
FIRST_STANDARD_PARALLEL = "N/A"
LINE_FIRST_PIXEL = 1
LINE_LAST_PIXEL = 12232
LINE_PROJECTION_OFFSET = -15833.625
MAP_PROJECTION_ROTATION = 0.0
MAP_PROJECTION_TYPE = SINUSOIDAL
MAP_RESOLUTION = 1185.49393024378 <pixel/degree>
MAP_SCALE = 0.05 <km/pixel>
MAXIMUM_LATITUDE = -13.3349
MINIMUM_LATITUDE = -23.7135
POSITIVE_LONGITUDE_DIRECTION = EAST
REFERENCE_LATITUDE = "N/A"
REFERENCE_LONGITUDE = "N/A"
SAMPLE_FIRST_PIXEL = 1
SAMPLE_LAST_PIXEL = 3210
SAMPLE_PROJECTION_OFFSET = 2061.625
SECOND_STANDARD_PARALLEL = "N/A"
WESTERNMOST_LONGITUDE = 176.2
END_OBJECT = IMAGE_MAP_PROJECTION

Those figures worked well (on the image from which they were taken, I just grabbed numbers from the blue .img - not an exact match, but it'll do) . I've dumped USGS Dem's and .ter's out of 3dem. 3ds max being it's usual tempremental self refused to import the DEM's - and while a free terragen importer plugin exists, it got as far as swallowing 3 gig of ram, and then I gave up

- Okino Polytrans happily imported the DEM, exported a 3ds, which 3ds max imported.

So - with your help, I've got a pipeline that works.

uncompress .bz2
ISIS3 vicar2isis
ISIS3 isis2raw
get lines and sample from .dt4
get lat/long extentts from a map-proj .img
import to 3dem
export dem
import to Polytrans
Export to 3ds max

As an example. I get a 2.5 million polygon model from 0072_0000 - which actually will optimize down to more like 100,000 with not much work. Going to work on getting some texture sorted from dropping the RGB over the ND, match it up and see how it renders out!

Many thanks for your help!

Doug

Posted: **02.01.2008, 21:58**

Thanks for refreshing my memory about the HRSC PDS library. I also used them for creating map files to mosaic a couple of dt4 files together and seemed to have forgotten.

As far as the PDS img files, I think they are level3 and are not calibrated to correct pixel alignment for rgb stacking. The level4 files are, which is probably where you got the dt4 files. So if you want to create a texture to go with the dem, you probably want to use the level4 files.
After importing them into ISIS, you can use the isis2std program and create grayscale png files. Then you stack them with your favorite image manipulation program.

Another thing about using the label data from the PDS HRSC img files, The min and max latitudes and longitudes don't always line up exactly between the dt4 files and the re4,gr4,bl4 and nd4 files. This can be corrected by using the crop program from ISIS and crop all the files so they all have exactly the same coordinates. You may need to do some math with pixels per degree, how many pixels were cropped out, and so on........
By using the gimp, I get an idea of what needs to be cropped by stacking the img file I want to use the label from with the dt4 file and see how many pixels have to go. Then I do the actual crop with ISIS and then I calculate the new min and max values.

You may also have minor errors in the min and max latitude and longitude in the dem if a label from one of the PDS HRSC img files had different min and max values from the dt4 file. see above.
Doing it right gets a little complicated

cartrite

Posted: **04.01.2008, 23:19**

Emily from TPS put me in touch with Greg Michael - who, inside of 24 hours, put an IDL program onto the HRSC website which does everything - all of it - in one hit.

It's at the bottom of this page
http://hrscview.fu-berlin.de/software.html

And I've put some of my early results here:
http://www.unmannedspaceflight.com/inde ... t&p=106812

Posted: **05.01.2008, 01:22**

I'm glad you are getting good results. Your models look very good. I will have to give this a try. By Tiffs I'm assuming they are Geo Tiffs?

Some of the results I came up with can be seen here. http://www.celestiaproject.net/forum/viewtopic ... 6252#96252
I also did a couple of short clips which are available on YouTube.
The full resolution version of one of them is on the Movie page at my website.
Here are the other clip links on YouTube.
http://www.youtube.com/watch?v=a-qEkKP1u-k
http://www.youtube.com/watch?v=kz5Lte6p_NE
http://www.youtube.com/watch?v=dssQhME4Ksc
The last 2 clips on You Tube were using MRO CTX textures but the DEMS were still from the HRSC camera on Mars Express.
cartrite

Celestia Forums

New Mars Image data available

New Mars Image data available