[ome-users] About OME-XML and image formats

[Jason Swedlow]

Hi Phillip--
On 14 May 2005, at 14:44, phillip ryan ganir wrote:

>      Good day!
>

To you too.  Is a sunny May morning in Scotland!  Yeah!

>      I'm studying image formats and planning to do a software that 
> converts
> image formats to image formats. I just want to know if my newly
> gained insights about OME-XML is true. In line with this are the 
> following
> questions:

Will try to help.

>
>      Is OME-XML another image file format like stk, tiff, jpeg, etc., 
> or is it
> not?

OK, there are many issues here.  The three file types you mention are 
very different.  STK supports multi-dimensional images, and is 
nominally proprietary (from MD/UIC)..  TIFF and JPEG are standards.  
Generally, they are used to store single images, and support various 
types of compression.  OME-XML differs in that it supports a 5D 
structure AND a

> In simpler terms, is a .ome file viewed as an image?


Yes, it can be, but see the next comment.

> If yes, is it
> necessary that such image viewer be internal to OME? Or can it be an 
> external
> viewer (just like a jpeg viewer for jpeg, a tiff viewer for tiff, 
> etc.)? If
> not, does OME have an image file format just like the ones mentioned?

OME XML is an open file format.  To our knowledge, their are 2 released 
commercial viewers reading and writing OME XML-- from Bitplane AG and 
Aplied Precision, LLC.  At least 3 other  companies have told us they 
have viewers that support OME XML under development.  Also, OME 
software can import OME XML into an OME database, where it is then 
accessible from one of our clients (http://openmicroscopy.org for more 
info).

>
>      Does every image format in an OME database, like the ones 
> mentioned can be
> converted to an OME-XML?

I am not sure if there is a bit of confusion here.  Just to be clear: 
we have a data model, that describes the relationships between various 
data types in  a light microscope image.  This is called the OME Data 
Model, and is described in a recent public (available at 
http://genomebiology.com/2005/6/5/R47).  This model is used as the 
basis of an OME database, built by OME server software (available at 
http://cvs.openmicroscopy.org.uk).  It is also the basis of the 
specification for the OME XML file (see 
http://openmicroscopy.org/api/xml/).

OME server software supports the export of images in a OME database 
into OME XML, so that data can be migrated from one site to another.

There is a major problem with the OME XML file format.  The binary data 
is encoded as base64, and then usually compressed.  Most viewers 
support file pointers, that allow the user to arbitrarily view a given 
timepoint, channel, or optical section (t, c, z, in OME parlance).  
After compression, it is not possible to access a random frame using a 
pointer, without decompressing the entire 5D image.  For lareg images, 
this creates a significant performance problem.  Thus, OME XML is best 
used as a data migration format. Clearly, we need an alternative that 
allows better performance.  We are working on this, and in 
collaboration with both academic and commercial partners will release 
an alternative format for data visualization.

> If this is the case, I think I can use this
> functionality for converting image formats to image formats (at least 
> to those
> OME supports) because each image format can be converted to a common
> format, so one can just view to different image format as two files of 
> the same
> format (in this case, OME's format). Am I getting it right?

Sounds good to me.

>
>      Does OME now support image format to image format translation 
> (perhaps in a
> way other than what I just wrote)?
>

Maybe.  OME server software lets you import data from a variety (and 
increasing) list of proprietary file formats.  This is stored with the 
OME database.  Data can then be exported to OME XML.  Whether this 
process is efficient for you depends on what you want.

>      About image formats in general, they are really in binary form, 
> right? And
> in this binary file, some have information, like metadata, embedded in 
> them,
> haven't they? Specifically, tiff and stk have this, right?

This really depends on the format.  Various methods are used.  For info 
on TIFF, see 
http://www.ee.cooper.edu/courses/course_pages/past_courses/EE458/TIFF/

>
>      I hope someone can find patience to answer my questions.:) I 
> would really
> appreciate future responses.

Hope this helps.

Cheers,

Jason

NOTE NEW EMAIL ADDRESS: jason at lifesci.dundee.ac.uk
**************************
MSI/WTB Complex
The University of Dundee
Dow Street
Dundee  DD1 5EH
United Kingdom

phone (01382) 345819
Intl phone:  44 1382 345819
FAX   (01382) 348072
email: jason at lifesci.dundee.ac.uk

Lab Page: http://www.dundee.ac.uk/lifesciences/swedlow/
Open Microscopy Environment: http://www.openmicroscopy.org
**************************
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 5215 bytes
Desc: not available
URL: <http://lists.openmicroscopy.org.uk/pipermail/ome-users/attachments/20050517/579e0ba7/attachment.bin>