On 03-Apr-04 Maura Edelweiss Monville wrote:
Is there a library or development environment equivalent to IDL (http://www.floating.co.uk/idl/) supported by Linux/SuSE or Linux in general ?
I'm particularly interested in BioMedical Imaging development. I'd rather discard any package that runs on top of Windows !!!
Hi Maura,
First of all, it seems from the IDL website that Linux is supported!
(They're coy about their pricing, though ... )
Next, looking at the functionality of IDL, I think (without checking
in detail) that a great deal of the IDL functionality of operations
on arrays can be found in octave: see
http://www.octave.org
Octave also has a well-developed set of libraries for the range of
signal analysis operations described for IDL. However, octave's
repertoire of "main-line" statistical analysis routines is somewhat
limited: a good bit of basic stuff is there, but not much of the
fancier procedures. Also, while octave can handle image analysis
up to a point, I'm not sure that it can match what IDL claims.
Octave has quite good data graphing capabilities but again, I think,
does not go as far as IDL seems to.
Finally, if you need access to really comprehensive statistical
analysis, you should have a look at R:
http://www.r-project.org
R also has very comprehensive array procedures (and lots of other
types of complex objects too). However, I'm not sure what you can
find there for image analysis (but this is mainly because I don't
have much to do with that area). There are hundreds of extra
libraries for R and it may well be that somewhere in there are
useful image analysis resources. Also, there are some specialist
groups using R who develop their own techniques, and there may
be an image-analysis group developing useful stuff. Whatever the
case may be on this front, there's no doubt that R has the capability
to support anything you might want to do with images -- it only needs
someone to implement it!
I hope these commments are useful.
Best wishes,
Ted.
--------------------------------------------------------------------
E-Mail: (Ted Harding)