Isis Support Center

Isis 3 Support :: RE: Incorrect behavior with reduce followed by cam2map (OPEN JA)

Author: janderson
Posted: Thu May 15, 2008 5:17 pm (GMT -7)
Topic Replies: 1

Hi,

Certainly looks like a problem to me. We will make the change to the code (AlphaCube) you pointed out and do some testing.

I will talk with some other folks around here about adding a forum for software development. I'm not sure how many people are actually coding in ISIS3 (other than the USGS folks) so it may not see much traffic. Its probably worth a shot though.

Jeff

Isis 3 Support :: RE: Cassini ISS to ISIS (OPEN JA)

Author: janderson
Posted: Thu May 15, 2008 2:50 pm (GMT -7)
Topic Replies: 1

Hi

I think you need to download the associated PDS lbl file which is detached from the image data. Then run the ciss2isis program giving FROM the lbl file.

Jeff

Isis 3 Support :: Cassini ISS to ISIS (OPEN JA)

Author: heckofajob
Subject: Cassini ISS to ISIS (OPEN JA)
Posted: Thu May 15, 2008 12:45 pm (GMT -7)
Topic Replies: 1

I have what appears to me to be an exceedingly dumb question, but after days of trying I can't figure it out.

I am a new ISIS3 user (usually I work with images from telescopes; much easier to understand), and I am trying to look at Cassini ISS images of Titan.

I download an image from the PDS, such as
W1477461455_6.IMG

I next try
ciss2isis from=W1477461455_6.IMG to=test.cub

But I just get:
**I/O ERROR** Unable to read PVL file [/home/heckofajob/isistest/W1477461455_6.IMG]
**PVL ERROR** Invalid PVL format
**PARSE ERROR** ASCII data expected but found unprintable (binary) data

If I instead try vicar2isis I can indeed get the image into an ISIS cube, but none of the label data are preserved so it doesn't do me much good.

I can read it into IDL with vicar_read.pro, which works nicely too for simply looking at the image, but doesn't help much with geometry.

Any hints anyone?

Isis 3 Support :: RE: ISIS3 CGI Script (OPEN TH)

Author: ollerery
Subject: Settingh Environment variable
Posted: Thu May 15, 2008 12:43 pm (GMT -7)
Topic Replies: 2

I have all the ISIS environment variables set for my web user but i still get a segmentation fault. I suppose that there could be some other path that i am forgetting to set. Here is what i currently have set for my regular user account which works, and my web user account which is giving me the segmentation fault.

This is what the Environment variables look like for my standard user account

Code:

MANPATH=/usr/share/man:/usr/local/share/man:/usr/X11/man
TERM_PROGRAM=iTerm.app
ISIS3DATA=/isis/isis3/isis/../data
SHELL=/bin/bash
TERM=rxvt
TMPDIR=/var/folders/LM/LMggH7QEFx8FzFa0vQFzuk+++TY/-Tmp-/
ISISROOT=/isis/isis3/isis
Apple_PubSub_Socket_Render=/tmp/launch-OGv0Kj/Render
ISIS3TESTDATA=/usgs/cpkgs/isis3/testData
USER=ryan
COMMAND_MODE=legacy
SSH_AUTH_SOCK=/tmp/launch-fia9oH/Listeners
__CF_USER_TEXT_ENCODING=0x1F9:0:0
PATH=/Library/PostgreSQL8/bin:/usr/bin:/bin:/usr/sbin:/sbin:/isis/isis3/isis/bin:/usr/local/bin:/usr/X11/bin:/usr/local/geojasper/
PWD=/Library/WebServer/CGI-Executables
HOME=/Users/ryan
SHLVL=2
DYLD_LIBRARY_PATH=/isis/isis3/isis/lib:/isis/isis3/isis/3rdParty/lib
LOGNAME=ryan
DISPLAY=/tmp/launch-NESMUy/:0
SECURITYSESSIONID=5b1a100
_=/usr/bin/env

This is what the Environment variables look like for my web user

Code:

SERVER_SIGNATURE=
ISIS3DATA=/isis/isis3/isis/../data
ISISROOT=/isis/isis3/isis
HTTP_KEEP_ALIVE=300
ISIS3TESTDATA=/usgs/cpkgs/isis3/testData
HTTP_USER_AGENT=Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9b5) Gecko/2008032619 Firefox/3.0b5
SERVER_PORT=80
HTTP_HOST=xxxx.xxxx.xxx.xxx
DOCUMENT_ROOT=/Library/WebServer/Documents
HTTP_ACCEPT_CHARSET=ISO-8859-1,utf-8;q=0.7,*;q=0.7
SCRIPT_FILENAME=/Library/WebServer/CGI-Executables/test-cgi
REQUEST_URI=/cgi-bin/test-cgi SCRIPT_NAME=/cgi-bin/test-cgi HTTP_CONNECTION=keep-alive
REMOTE_PORT=xxxxxx
PATH=/usr/bin:/bin:/usr/sbin:/sbin:/isis/isis3/isis/bin
PWD=/Library/WebServer/CGI-Executables
SERVER_ADMIN=you@example.com
HTTP_ACCEPT_LANGUAGE=en-us,en;q=0.5
HTTP_ACCEPT=text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 REMOTE_ADDR=xxx.xx.xxxx.xxxx
SHLVL=1
SERVER_NAME=xxxx.xxxx.xxx.xxx
DYLD_LIBRARY_PATH=/isis/isis3/isis/lib:/isis/isis3/isis/3rdParty/lib SERVER_SOFTWARE=Apache/2.2.8 (Unix) mod_ssl/2.2.8 OpenSSL/0.9.7l DAV/2 PHP/5.2.5 QUERY_STRING=
SERVER_ADDR=xxx.xxx.x.xxx
GATEWAY_INTERFACE=CGI/1.1 SERVER_PROTOCOL=HTTP/1.1 HTTP_CACHE_CONTROL=max-age=0
HTTP_ACCEPT_ENCODING=gzip,deflate
REQUEST_METHOD=GET _=/usr/bin/ENV

Isis 3 Support :: Hirise stretch issue

Author: cip
Subject: Hirise stretch issue
Posted: Thu May 15, 2008 11:38 am (GMT -7)
Topic Replies: 0

Hi,
Trying to save time I downloaded the files from Hirise site (jpeg2000 format). In order to save some space and use them into some GIS soft I tried to convert them from the declared 16 bit format to 8 bit raw format with envi and project to different map. The result is what you see in the screen capture atached (the majority of high end DN turn to 255 maximum in 8bit). Tried to import it to ISIS3 with raw2isis, use the stretch from ISIS2raw to a 8 bit and the same result.
Is there a special value in the data that mess the statistics? Actualy when I do the statistics on this particular image it says that the values range from 0-1023 but the last reported value in the list is 1020. See 2nd atachment.
Any idea how to avoid this? And is not the only one that behaves this way.

Isis 2 Support :: RE: qview bus error

Author: schenk
Posted: Wed May 14, 2008 4:08 pm (GMT -7)
Topic Replies: 3

forgot.

10.5.2 system....

Isis 2 Support :: Cassini ISS to ISIS conversion

Author: drallen1
Subject: Cassini ISS to ISIS conversion
Posted: Wed May 14, 2008 3:22 pm (GMT -7)
Topic Replies: 0

I am currently working on some raw Cassini ISS images and need to convert them to ISIS format. I tried raw2isis, and it gave me the following error message:

[TAE-TMISPAR] Missing parameters: ITYPE, NS, NL

I believe the NS and NL are 1024 and the ITYPE is 2. I tried that, and the resulting .cub file doesn't look right. I have some images that were previously processed using vicar2isis, but I understand that doesn't exist in Isis 2. Any help or suggestions would be appreciated.

Isis 3 Support :: RE: ISIS3 CGI Script (OPEN TH)

Author: thare
Posted: Wed May 14, 2008 8:56 am (GMT -7)
Topic Replies: 2

I don't have a solution but when and how unix reset environments has always been tricky for me to understand.

I was given a hint that maybe you need set up the enivronment correctly for the "webuser" then the cgi script whould run under that. Hopefully the webusers isn't root but a limited account for apache or other server software.

This will probably take some looking into on how to set up web cgi script environment properly.

Trent

GIS and Isis :: RE: Projections used in "daisy" to make globes

Author: thare
Posted: Tue May 13, 2008 8:18 pm (GMT -7)
Topic Replies: 8

To run the PERL script, make_gores.pl, you just need the file to be PPM so I don't think you need the PPM library. But maybe you are just looking for Netpbm anyway. To save out a PPM file from other formats you can use the freeware imagemagick (for most systems) or irfanview (windows). If you want ppmglobe then you do need netbpm but it looks like it is included on most systems including cygwin for Windows. older list of OSs: http://www.math.utah.edu/~beebe/unix/p/ppmglobe.html

BTW, if you are stuck on Windows and can't run PERL scripts you can get ActivePerl for Windows. Once installed, you can check out perl2exe to create an executable from a PERL script.

-Trent

GIS and Isis :: RE: Projections used in "daisy" to make globes

Author: globuli
Posted: Tue May 13, 2008 6:48 pm (GMT -7)
Topic Replies: 8

netpbm[1] is open source. you can get it using your packet manager, or typing apt-get netpbm in terminal if you are on linux. but: chances are, that you will still miss ppmglobe since it is not included in the debian package[2] of netpbm. but you can download the source and just compile ppmglobe only and put it to the rest of the netpbm package, that works for me.

[1] http://netpbm.sourceforge.net
[2] http://netpbm.alioth.debian.org

GIS and Isis :: RE: Projections used in "daisy" to make globes

Author: 0adrian
Posted: Tue May 13, 2008 6:38 pm (GMT -7)
Topic Replies: 8

Hi all, I'm new to this forum.

I just wonder is Netpbm freeware? How to get it?

Thanks,

Adrian
_________________
Software Library

Isis 2 Support :: RE: qview bus error

Author: schenk
Posted: Tue May 13, 2008 4:57 pm (GMT -7)
Topic Replies: 3

most of the rest of ISIS2 works fine, although i havent tried everything.

paul

Isis 3 Support :: Incorrect behavior with reduce followed by cam2map (OPEN JA)

Author: mbroxton
Subject: Incorrect behavior with reduce followed by cam2map (OPEN JA)
Posted: Tue May 13, 2008 2:10 pm (GMT -7)
Topic Replies: 1

Hi there,

I'm using ISIS 3.1.15 on an Intel Mac running OSX 10.5.2, and I've encountered some strange behavior when I try to subsample an image using the 'reduce' command and then map project it using 'cam2map'.

Starting with a raw Clemenetine image I imported into ISIS, I run:

Code:

> reduce from=lua0825f.cub to=lua0825f-reduce.cub lscale=2 sscale=2
> cam2map from=lua0825f-reduce.cub to=lua0825f-reduce-map.cub

as well as

Code:

> cam2map from=lua0825f.cub to=lua0825f-map.cub
> reduce from=lua0825f-map.cub to=lua0825f-map-reduce.cub lscale=2 sscale=2

Then, using qview or 'camrange' tool, I check to see if the lat/lon extents of the reduced image are still the same as the original.

Here's are the original lat/lon extents:

Code:

> camrange from=lua0825f.cub
Group = UniversalGroundRange
....
MinimumLatitude = -36.712850392335
MaximumLatitude = -34.731334937095
MinimumLongitude = 97.010056321466
MaximumLongitude = 100.27835504082
....

And here are the extents for the reduced, then map projected image. As you can see, the extents have been incorrectly(?) reduced by a factor of two.

Code:

> camrange from=lua0825f-reduce-map.cub

Group = UniversalGroundRange
...
MinimumLatitude = -36.712850392335
MaximumLatitude = -35.724705674756
MinimumLongitude = 98.645182040178
MaximumLongitude = 100.27835504082
...

And here are the extents if you map project first, then reduce the image. Here the lat lon extents are almost identical to the original image. (I'm assuming the difference is probably due to a 0.5 pixel offset...)

Code:

> camrange from=lua0825f-map-reduce.cub

Group = UniversalGroundRange
...
MinimumLatitude = -36.717890136024
MaximumLatitude = -34.729458400171
MinimumLongitude = 97.004622729826
MaximumLongitude = 100.30077912936
...

I dug around a bit and I found what looks to be the relevant bit of code on lines 83-104 of src/base/apps/reduce/reduce.cpp. I've added some of my own comments inline with the code below to point out a few things that have me a little confused.

Code:

try {
PvlGroup &mapgroup = cube.Label()->FindGroup("Mapping", Pvl::Traverse);

if(sscale != lscale) {
//
// This if clause has me a little confused. Am I correct in
// interpreting that if the scale factors are non-isotropic,
// 'reduce' will blow away the map projection information
// entirely?
//
opack->DeleteGroup("Mapping");
}
else {
//
// This is the code that correctly changes the scale factor
// when your run 'reduce' on a map projected image. However,
// I was wondering why it only changes the "PixelResolution" tag,
// and not also the "Scale" tag. (My map projected Clementine
// image had both tags, and only one was correctly updated.)
//
double pixres = mapgroup["PixelResolution"];
mapgroup["PixelResolution"] = pixres * sscale;
opack->PutGroup(mapgroup);
}
}
catch(iException &e) {
e.Clear();
//
// I think that the following line is the root of the problem.
// If I understand AlphaCubes correctly, I would have thought
// that the "AlphaEndingSample" and "AlphaEndingLine" should
// be left unchanged by 'reduce'; it should only affect the "BetaLines"
// and "BetaSamples" tags in the AlphaCube group.
// it should read as:
//
// AlphaCube alpha(cube.Samples(), cube.Lines(),
// opack->Samples(), opack->Lines(),
// 0.5, 0.5, cube.Samples()+0.5, cube.Lines()+0.5);
//
AlphaCube alpha(cube.Samples(), cube.Lines(),
opack->Samples(), opack->Lines(),
0.5, 0.5, ons+0.5, onl+0.5);
opack->UpdateAlphaCube(alpha);
}

I also wanted to point to this post, which looks to be closely related to this problem: http://isis.astrogeology.usgs.gov/IsisSupport/viewtopic.php?t=1510&sid=85f9ef28fa4c22dd38870b7f855185b6

I think that's probably enough information to get the thread started on this issue. Thanks in advance for taking the time to look at it.

One other quick related question while I'm posting -- I was thinking that it would be great if you added a new forum within the ISIS Support Center for programming and under-the-hood types of questions such as this. Something like "ISIS Development" would be a handy place for folks like me to share our experiences in compiling our own code against the ISIS libraries and building ISIS ourselves.

Thanks again!

Isis 3 Support :: ISIS3 CGI Script (OPEN TH)

Author: ollerery
Subject: ISIS3 CGI Script (OPEN TH)
Posted: Tue May 13, 2008 12:58 pm (GMT -7)
Topic Replies: 2

I am trying to use ISIS within a perl CGI script but when the ISIS call is made I get a segmentation fault. I am using a shell script to set up all my ISIS environment variables and then to make the ISIS call, which is called by the CGI script. When i execute the shell script from the command line it works fine. But when i access the cgi script from a web browser the print lines will execute but not the ISIS commands.

Here is my cgi script

Code:

#!/usr/bin/perl

print "Content-type: text/html\n\n";
print "<html><head>";
print "<title>CGI Test</title>";
print "</head>";
print "<body><h1>PERL Header</h1>";

my $listing = `./vtest.sh`;
my @output = split(/\n/,$listing);

print "spiceinit output: $listing <br>\n\n";

foreach $out (@output) {
print "-> $out <br>\n";
}

print "</body></html>";

Here is the shell script

Code:

#! /bin/bash

ISISROOT=/isis/isis3/isis
export ISISROOT

. $ISISROOT/scripts/isis3Startup.sh

spiceinit from=/test/PSP_001503_1645_RED9.histitch.cub

echo "End Perl script"

Here is the error that i am getting in my apache error_log

Code:

./vtest.sh: line 9: 5017 Segmentation fault isis2std from=/test/PSP_001503_1645_RED9.histitch.cub

Isis 2 Support :: RE: qview bus error

Author: jani
Posted: Tue May 13, 2008 12:56 pm (GMT -7)
Topic Replies: 3

This may be a problem with Leopard. Nice test case - two computers identical in all but OS: The one with Tiger (Mac OS 10.4.11) runs ISIS2, but the one with Leopard (10.5) gives me a bus error.

Are there any plans to compile ISIS2 on Leopard? If not, we've got to wait to upgrade, but will probably be forced by computer society to cave eventually...

Thanks -
Jani