| Re: Array resize with arbitrary arithmetic [message #53010 is a reply to message #52942] |
Tue, 13 March 2007 20:26  |
Dick Jackson
Messages: 347
Registered: August 1998
|
Senior Member |
|
|
Hi Radha,
"David Fanning" <news@dfanning.com> wrote in message
news:MPG.205f7ed4a7dbb8da989ed1@news.frii.com...
> David Fanning writes:
>
>>> image=read_bmp('9 March 2007\775u.bmp') ; 640x240 array
>>> temp = Rebin(image, 2, 320, 240)
>>> temp = Total(Temporary(temp), 1)
> [...]
> I *think* the answer to the original question, which
> looks to me like we want to multiply the first pixel
> by 256 and add the second, adjacent pixel to it (does
> that seem weird to you!?) is this:
>
> image=read_bmp('9 March 2007\775u.bmp') ; 640x240 array
> temp = Rebin(image, 2, 320, 240) ; Adjacent pixels in cols
> temp[0,*,*] = temp[0,*,*] * 256 ; Multiply 1st col by 256.
> temp = Total(temp,1) ; Add columns together.
That's not bad, but if I'm right, this arithmetic isn't so arbitrary! We're just
turning each pair of bytes into an unsigned short integer, so this one line
should do the whole thing:
comp_image = UInt(image, 0, 320, 240)
... except if you're on a "little-endian" machine, you'll need to swap the
bytes. This statement will work to fix it if needed:
Swap_Endian_InPlace, comp_image, /Swap_If_Little_Endian
Now, I don't have your data, but this test should prove the point. I tested it
on Intel (little-endian) and works fine, can someone double-check it on
big-endian hardware, please?
IDL> image=bindgen(6,2)*21B
IDL> print,image
0 21 42 63 84 105
126 147 168 189 210 231
IDL> comp_image = UInt(image, 0, 3, 2)
IDL> Swap_Endian_InPlace, comp_image, /Swap_If_Little_Endian
IDL> print,comp_image
21 10815 21609
32403 43197 53991
If you really need the result as Long, then *after* all this, do
comp_image = Long(comp_image)
Hope this helps!
--
Cheers,
-Dick
--
Dick Jackson Software Consulting http://www.d-jackson.com
Victoria, BC, Canada +1-250-220-6117 dick@d-jackson.com
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
