| Re: trying to export pixel data from .dat files, based on coordinate loc [message #71697 is a reply to message #71616] |
Mon, 12 July 2010 14:39   |
Maxwell Peck
Messages: 61
Registered: February 2010
|
Member |
|
|
On Jul 13, 7:29 am, Snow53 <jennifer_wa...@hotmail.com> wrote:
> On Jul 12, 3:12 pm, Maxwell Peck <maxjp...@gmail.com> wrote:
>
>
>
>> On Jul 13, 6:37 am, Snow53 <jennifer_wa...@hotmail.com> wrote:
>
>>> Thanks to all who gave great advice. I almost have this up and
>>> running, but have a few more questions.
>
>>> 1. At the moment, I have tested this code based on two files that each
>>> have their own .hdr. For the real thing, I want to create only
>>> one .hdr that can be used for all (200+) files since they have the
>>> same dims, data type, etc. How can I modify this code to look for that
>>> one .hdr file and use information from that file when looping through
>>> each file in the folder?
>
>>> 2. The ENVI Available Bands GUI pops up when I run this. Is this
>>> supposed to happen? I read that envi_open_file was non-interactive in
>>> idl (with batch mode).
>
>>> 3. The way the code currently reads, it will output twice when I run
>>> it but only giving me pixel data from the first file read under
>>> 'file'. I think that I need to write a loop to specify to read from
>>> the 'file' list one at a time, go through the code, close that file,
>>> and then start with the next. I'm not sure, though, how to write
>>> this, and would appreciate advice.
>
>>> 4. I did notice that I'm not getting back the correct sample/line
>>> pixel file locations from my input map locations (x,y). They seem to
>>> be one pixel off. Has anyone else had this happen?
>
>>> The code is shown below. Thanks again!
>
>>> Goal: Extract pixel data based on input coordinate location for each
>>> file (ENVI binary)
>>> ; within a specified folder location. Export this data to a .csv file.
>
>>> pro extractdata4
>
>>> ;define path
>>> cd, 'X:\MERRA\HDF_Output_Lena\test\'
>
>>> ;open envi files within given folder
>>> file_array=file_search('*.dat', count=num_file)
>
>>> for i=0, num_file-1 do begin
>>> file=file_array
>>> endfor
>>> print, num_file
>>> print, file
>
>>> ;read ENVI binary files
>
>>> envi_open_file , file, r_fid=fid
>
>>> ;convert x,y map coordinates to corresponding pixel coordinates. note
>>> that xmap and ymap can be single values or arrays if
>>> ;needed to extract info for multiple pixels.
>>> XMap=[109.55551335]
>>> YMap=[79.25]
>
>>> ENVI_CONVERT_FILE_COORDINATES, fid,XF, YF, XMap, YMap
>
>>> XF_out=Round(XF)
>>> YF_out=Round(YF)
>>> print, 'x pix' ,XF_out
>>> print, 'y pix', YF_out
>
>>> ;specify the data dims for the pixels who's info you want to extract.
>>> pos specifies which band(s) you want to extract from.
>>> ;for example, if I have 4 bands and I only want to extract from bands
>>> 1 and 4, pos would be [0,3]. Then extract for these pixels/bands.
>>> dims=[-1, XF_out, XF_out, YF_out, YF_out]
>>> pos=[0]
>>> pixdata = ENVI_GET_DATA(fid=fid, dims=dims, pos=pos)
>
>>> print, pixdata
>>> ;open text file to write data to
>>> OPENU, U, 'pixel_value.csv', /get_lun, /append
>>> ;write data
>>> printf, U, pixdata
>>> ;close LUN
>>> close, U
>
>>> end
>
>> 1. It can be done just using IDL to read in the images but it will
>> probably be easier for you just to duplicate the header file and keep
>> them in ENVI format. (Unless you wanted to completely rewrite the
>> program)
>
>> 2. You probably want ,/NO_REALISE in your envi_open_file.
>
>> 3. Yes, use ENVI_FILE_MNG to close the file each time after the map
>> conversion.
>
>> 4. Is it one pixel off or half a pixel off ?
>
>> Max- Hide quoted text -
>
>> - Show quoted text -
>
> Thanks Max.
>
> I just checked and it is one pixel off, not 0.5.
Ok, im assuming you're using ENVI to check. Envi starts at 1,1 in the
top left whereas the conversion i think uses 0,0 . I don't have envi
here to check at the moment so I'd take a look yourself.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
