[McIDAS #WWL-217735]: pixel shift in Florida Satellite Water Vapor Imagery

Hi Daphne,

I apologize for the silence today... there is a Unidata User's Committee meeting
here in Boulder today and tomorrow, so my day is pretty much booked except for
breaks.

re:
> Whenever you get a chance, can you share the procedure used to shift the 3.9
> image and create the fog image?

Basically, what I did in McIDAS-X was:

- load a 3.9 um image in frame 1 so that the upper left corner of the image
  was loaded in the upper left corner of my display and blow up the display
  by a factor of 3

  The image was displayed using the McIDAS IMGDISP command.

- load a 10.7 um image in frame 2 so that the upper left corner of the image
  was loaded in the upper left corner of my display and blow up the display
  by a factor of 3

  The image was displayed using the McIDAS IMGDISP command.

- loop frames 1 and 2 so I could verify that there is indeed a displacement

  McIDAS-X:

  LB 1 1

- adjust image directory word 6 (UPPERLEFTLINE: Image line coordinate of area
  line 0, elem. 0) a pixel at a time to see what shift would result in the
  best correspondence between the 3.9 um and 10.7 um images

  This required that I reload the display of the 3.9 um image after adjusting
  image directory word 6 and then animating the 3.9 and 10.7 um images.

  Word 6 in the 3.9 um image header was adjusted using the McIDAS-X command
  LWU POKE

- what is needed in McIDAS-X at this point is to create new 3.9 um and 10.7 um
  images so that all pixels in the 3.9 um image are co-located with 
corresponding
  pixels in the 10.7 um image.  This must be done in McIDAS-X since the 
navigation
  and spatial extents of images which will be used in combinations like that of
  a fog product must be identical.

  These would be accomplished using the McIDAS-X IMGCOPY command for each
  band.

- the final step is to use the McIDAS-X IMGOPER command to combine the adjusted
  3.9 and 10.7 um images to create a fog product

  This is accomplished using the McIDAS-X IMGOPER command.

I intend to put together a Unix shell script that will do all of the steps
above to demonstrate the process I was using, but I have not had time to
get to the job.  I apologize for this!

re:
> Also, did you use McIdas V or X to do this? We are currently working with
> McIdas V.

I don't know of an easy way to change image header values in McV/IDV (there
may be one, but I am not familiar with it).  Using McIDAS-X to make the
header change is very simple, however, and I think that after making the
change McV/IDV could be used to create the fog product composite directly.

The thing that must be done before bulk processing of images is to verify
that the amount of pixel shift needed is the same for all of the images
in your dataset.  Unfortunately, this will need to be done by-hand and
with visual examination of the registration through the first two steps that
I outlined above.

Question:

- how many images pairs (3.9 and 10.7 um images from the same time) are
  in your dataset?

- are the areal coverages the exact same for all of the images in your
  dataset?

  If yes, there is a good chance that the needed pixel shifts will be
  the same thus eliminating the very tedious step of comparing each
  of the 3.9/10.7 um image pairs.

re:
> Let us know, we greatly appreciate your help!

Is anyone there familiar with McIDAS-X?

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: WWL-217735
Department: Support McIDAS
Priority: Normal
Status: Closed