[ome-devel] Rotation and superposition of images with OME-TIFF?

Andrew Patterson ajpatterson at lifesci.dundee.ac.uk
Tue May 29 13:53:38 BST 2012


Hello Éric,

At the moment the core OME-TIFF has no concept of rotation. We have put forward a proposal for storing extra dimensions including Angle (mainly to allow SPIM data to be represented).

https://www.openmicroscopy.org/site/support/file-formats/working-with-ome-xml/6d-7d-and-8d-storage

This might be a bit heavy handed for what you need. It does indicate the general way of extending the model for storing additional data. 

By placing you data in Annotations with a defined namespace your software can correctly interpret the data. Other applications will just see them as standard OME-TIFF files. You could perhaps attach your information as annotations to each Image node.

On the final point, there is no way of saying an individual image is just a thumbnail. But you could add an annotation to that affect, it would just be meaning full to your software though. Other applications would see it as a full image. There are several other file formats the bio-formats reads that do produce similar data with thumbnail images mixed in with the full data. 

Hope this helps,

Andrew

On 22 May 2012, at 10:11, Éric Piel wrote:

> Hello,
> 
> I'd like to extend our software to export acquired image in OME-TIFF, and I have a couple of questions related to our specific needs.
> 
> First, let me introduce slightly the context. The hardware is a microscope we are developing that acquire images of the same sample both in epi-fluorescence and in SEM [1]. After calibration, the hardware has the two pictures approximately aligned (translation, rotation) but via software we can provide much finer precision. We'd like to save one acquisition (i.e., two images) into one OME-TIFF file. Here come my questions:
> 
> * I think I've found out how to indicate density (mathematically, scale) and position (mathematically, translation) of the two images, but I haven't found any way in OME-TIFF to indicate a rotation. Have I overlooked an attribute? Or what would be your recommendations to extend OME-TIFF to support such information? In GeoTIFF, there is a nice "coordinate-transformation matrix" but I haven't found anything fully similar in OME-TIFF.
> 
> * To display the two pictures together, we apply some kind of "blending filter" (and the user can change some parameters of this filter). As OME-TIFF was not developed with this in mind, it's obvious there is no support for such information. Any recommendation on how we'd best extend OME-TIFF for containing this information?
> 
> * Last question (a bit "cherry on top"). Is there a explicit way to record also in the file a "WYSIWYG" picture of the acquisition image as the user saw it in the software? In our case, this picture can be a superposition of several images cropped, rotated, blended, etc. So it'd be an image which is quite away from raw data, but more like how the user want to see the data. I'm thinking of putting this image as first image of the OME-TIFF (so that basic programs just display this one) and make sure to describe it in the OME attributes as different from the rest. Or is there a explicit way in OME-TIFF to say it's basically just a "big thumbnail"?
> 
> Thanks in advance for the insights you can provide!
> Éric Piel
> 
> 
> [1] http://delmic.com/products/life-sciences/
> _______________________________________________
> ome-devel mailing list
> ome-devel at lists.openmicroscopy.org.uk
> http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-devel



More information about the ome-devel mailing list