X
PrevPrev Go to previous topic
NextNext Go to next topic
Last Post 27 Jun 2015 09:24 AM by  anon
Trouble with TEXTURE_COOCCUR_DOIT
 10 Replies
Sort:
You are not authorized to post a reply.
Author Messages

anon



New Member


Posts:29
New Member


--
27 Jun 2015 09:24 AM
    Hi all, When I run this command with all the set parameters (based on the syntax at the top of http://www.exelisvis.com/..._COOCCUR_DOIT.html), including setting /G_LEVELS=64, I get an error reporting a syntax error when I try to compile my module, but when I take out the /G_LEVELS option, the script runs without any syntax errors. How come this is so? Is this option not supported in ENVI Classic 5.0, as the directions are for ENVI 5.2 SP 1? Thanks!

    MariM



    Veteran Member


    Posts:2396
    Veteran Member


    --
    29 Jun 2015 06:46 AM
    Try setting the parameter without the '/ ' which usually is just a toggle for on/off for a parameter. Using g_levels, you want to set it to a value. For example: g_levels=64

    Deleted User



    New Member


    Posts:29
    New Member


    --
    01 Jul 2015 10:18 AM
    Ok - so just to verify, I should slip in ", g_levels=64" instead of ", /g_levels=64"?

    Deleted User



    Basic Member


    Posts:228
    Basic Member


    --
    01 Jul 2015 10:58 AM
    No, that would be /g_levels instead of g_levels=64. I hope that works for you. - Peg Exelis VIS

    Deleted User



    New Member


    Posts:29
    New Member


    --
    02 Jul 2015 09:32 AM
    So I just put in /g_levels to turn it on? How do I specify 64 then?

    Deleted User



    New Member


    Posts:29
    New Member


    --
    06 Jul 2015 08:20 AM
    Thanks for helping with figuring out the g_levels switch. It worked when I just put in "g_levels=64" without the slash. It's helped me move along a lot on my project. Once I got past that, I noticed that once my raster data passed through the Texture_CoOccur_DoIt procedure, I just got 70 back garbled blank bands, when the input data has only 5 and when using the GUI in ENVI returns five layers bands 5 bands back. What could be wrong?

    MariM



    Veteran Member


    Posts:2396
    Veteran Member


    --
    06 Jul 2015 09:49 AM
    Yes, the G_LEVELS keyword is an assignment to a scalar value up to 256. If you process your 70 band image in the application using quantization of 64, do you get the same results as your code? If you do not use quantization, do you get a result?

    Deleted User



    New Member


    Posts:29
    New Member


    --
    10 Jul 2015 01:31 PM
    When I run it without the IDL script with 64 gray scale quantization, 5 bands go in, and 5 bands go out. Without the script nor gray quantization, the same comes out. When I did it without the gray scale quantization, I still had the same issue.

    Deleted User



    New Member


    Posts:29
    New Member


    --
    11 Jul 2015 02:00 PM
    It turned out that I was using the wrong dimensions and band names for this do_it command. Removing the g_levels option didn't change anything. The only issue I have now is that when I run the script, I get an error: "File: No Carriage returns found. Unable to read the file as an ASCII file" Then I get this error: "File: pathname to input LAS file Projection: 1.000 Problem: An unknown map projection is referenced in the map projection. Unable to use". This happens earlier in the script as it reads the projection information and dimensions as it converts .LAS data into ENVI data to go through the co-occurence measures. Is this likely because I'm using Windows and it is struggling with UNIX text files?

    MariM



    Veteran Member


    Posts:2396
    Veteran Member


    --
    13 Jul 2015 12:09 PM
    If the file is in LAS format, it should not matter. I assume you are using ENVI_CONVERT_LIDAR_DATA_DOIT to open the file? Did you set up the omap? Are you able to successfully import the file using the interface?

    Deleted User



    New Member


    Posts:29
    New Member


    --
    20 Jul 2015 02:32 PM
    I talked with the source of the files and they did something weird with the map projection data. Once I reprojected the data, it worked. Thanks for your standing by while I figured this out!
    You are not authorized to post a reply.