[ImageJ-devel] Save/write image with SCIFIO/BF in python script

Mark Hiner hiner at wisc.edu
Tue Jan 13 15:20:36 CST 2015


Hi Hadrien,

 Unsurprisingly, cropping and writing OME-TIFFs was quite broken.

 It's working for me now, as of the latest SCIFIO[1] and SCIFIO-OME-XML[2].
Note that these libraries have updated dependencies compared to what's
currently on Fiji, so you can't just drop these .jars into your Fiji
installation if you want to test locally; you need to update the
dependencies as well. The easiest way to do this would be to just clone
Imagej.git[3] and install it into your Fiji.app directory by running:

mvn -Dimagej.app.directory=/path/to/Fiji.app/ -Ddelete.other.versions=true

We'll push these changes up to Fiji this week, and I will find or write
better instructions for local testing.

Hope this helps. Let us know if you have any questions/problems. Thanks
again for finding this issue!

Best,
- Mark

[1] https://github.com/scifio/scifio/releases/tag/scifio-0.19.0
[2]
https://github.com/scifio/scifio-ome-xml/releases/tag/scifio-ome-xml-0.12.0
[3] https://github.com/imagej/imagej

On Sat, Jan 10, 2015 at 5:13 AM, Hadrien Mary <hadrien.mary at gmail.com>
wrote:

> Ok. Let me know when you want me to test your changes. I also tested
> fix-writing branch and I didn't notice any changes.
>
> --
> Hadrien Mary
>
> Ph.D student in Biology
> Tournier-Gachet Team
> CNRS - LBCMCP - UMR 5088
>
> Université de Toulouse - Bât. 4R3B1
> 118, route de Narbonne - 31062 Toulouse
>
>
> On Fri, Jan 9, 2015 at 10:14 PM, Mark Hiner <hiner at wisc.edu> wrote:
> > Hi Hadrien,
> >
> > I've identified at least two bugs that could be affecting you here.
> >
> > First of all, there was a bug in the OME-TIFF format writing out
> > multi-dimensional images. That's fixed on a branch but not uploaded to
> Fiji
> > yet [1].
> >
> > Second, it seems like cropping is restricting pixel reads to the correct
> > region but then doesn't change the image size.
> >
> > I'm hoping when I fix the second issue writing cropped regions will be
> > fixed.
> >
> > Thanks for the great feedback!
> > - Mark
> >
> > [1] https://github.com/scifio/scifio-ome-xml/tree/fix-writing
> >
> > On Fri, Jan 9, 2015 at 11:49 AM, Hadrien Mary <hadrien.mary at gmail.com>
> > wrote:
> >>
> >> I also confirm the same behaviour when replacing original file by OME
> >> sample data :
> >>
> http://www.openmicroscopy.org/Schemas/Samples/2013-06/bioformats-artificial/multi-channel-4D-series.ome.tif.zip
> .
> >>
> >> --
> >> Hadrien Mary
> >>
> >> Ph.D student in Biology
> >> Tournier-Gachet Team
> >> CNRS - LBCMCP - UMR 5088
> >>
> >> Université de Toulouse - Bât. 4R3B1
> >> 118, route de Narbonne - 31062 Toulouse
> >>
> >>
> >> On Fri, Jan 9, 2015 at 6:46 PM, Hadrien Mary <hadrien.mary at gmail.com>
> >> wrote:
> >> > After some investigations I think there is a bug with ImgSaver. In the
> >> > following script, X, Y crop is fine but axes informations (T and Z)
> >> > are lost and all frames appears on the same axes. Now if I replace
> >> > "target = "/home/hadim/cropped.tif" by "target =
> >> > "/home/hadim/cropped.ome.tif" on line 9. X, Y crop fails, cropped
> >> > image has the same dimensions than original however axes informations
> >> > (T and Z) are ok. Note that in cropped.ome.tif, pixel values are
> >> > totally incoherent which is not the case in cropped.tif.
> >> >
> >> > Files can be found here :
> >> >
> >> >
> https://arwen.hadim.fr/public.php?service=files&t=71aa17336dcccb3437d9a3dff789d2c0
> >> >
> >> > Archive contains cropped.tif, cropped.ome.tif and original.ome.tif.
> >> >
> >> > The script:
> >> >
> >> > ---------------------------------------------------
> >> > from io.scif import SCIFIO
> >> > from io.scif.config import SCIFIOConfig
> >> > from io.scif.img import ImageRegion
> >> > from io.scif.img import ImgOpener
> >> > from io.scif.img import ImgSaver
> >> > from net.imglib2.meta import Axes
> >> >
> >> > fname = "/home/hadim/original.ome.tif"
> >> > target = "/home/hadim/cropped.tif"
> >> >
> >> > axes = [Axes.X, Axes.Y]
> >> > ranges = ["%i-%i" % (0, 15), "%i-%i" % (0, 25)]
> >> > config = SCIFIOConfig()
> >> > config.imgOpenerSetRegion(ImageRegion(axes, ranges))
> >> >
> >> > opener = ImgOpener()
> >> > imps = opener.openImgs(fname, config)
> >> > imp = imps[0]
> >> >
> >> > saver = ImgSaver()
> >> > saver.saveImg(target, imp)
> >> >
> >> > print('Done')
> >> > ---------------------------------------------------
> >> >
> >> >
> >> > --
> >> > Hadrien Mary
> >> >
> >> >
> >> > On Fri, Jan 9, 2015 at 12:25 AM, Hadrien Mary <hadrien.mary at gmail.com
> >
> >> > wrote:
> >> >> Thanks for the tip (reminder to me: always look for example in
> tests!).
> >> >>
> >> >> Unfortunately crop does not work... I will try to do more tests
> >> >> tomorrow or wait for you or someone else to have a look.
> >> >>
> >> >> Updated code:
> >> >>
> >> >> from ij import IJ
> >> >> from ij import ImagePlus
> >> >>
> >> >> from io.scif import SCIFIO
> >> >> from io.scif.config import SCIFIOConfig
> >> >> from io.scif.img import ImageRegion
> >> >> from io.scif.img import ImgOpener
> >> >> from io.scif.img import ImgSaver
> >> >> from net.imglib2.meta import Axes
> >> >>
> >> >> fname = "/home/hadim/original.ome.tif"
> >> >> target = "/home/hadim/cropped.ome.tif"
> >> >>
> >> >> axes = [Axes.X, Axes.Y]
> >> >> ranges = ["%i-%i" % (2, 15), "%i-%i" % (2, 25)]
> >> >> config = SCIFIOConfig()
> >> >> config.imgOpenerSetRegion(ImageRegion(axes, ranges))
> >> >>
> >> >> opener = ImgOpener()
> >> >> imps = opener.openImgs(fname, config)
> >> >> imp = imps[0]
> >> >>
> >> >> saver = ImgSaver()
> >> >> saver.saveImg(target, imp)
> >> >>
> >> >> print('Done')
> >> >>
> >> >> --
> >> >> Hadrien Mary
> >> >>
> >> >>
> >> >> On Fri, Jan 9, 2015 at 12:09 AM, Curtis Rueden <ctrueden at wisc.edu>
> >> >> wrote:
> >> >>> Hi Hadrien,
> >> >>>
> >> >>>> I would like to be able to crop a region in X and Y while
> >> >>>> keeping any other dimensions which could exist (C, Z, T).
> >> >>>
> >> >>> Here's some Java:
> >> >>>
> >> >>>     int minX = 128, maxX = 255, minY = 128, minY = 255;
> >> >>>     AxisType[] axes = { Axes.X, Axes.Y };
> >> >>>     String[] ranges = { minX + "-" + maxX, minY + "-" + maxY };
> >> >>>     config.imgOpenerSetRegion(new ImageRegion(axes, ranges));
> >> >>>
> >> >>> Adapted from here:
> >> >>>
> >> >>>
> https://github.com/scifio/scifio/blob/scifio-0.18.0/src/test/java/io/scif/img/utests/ImgOpenerTest.java#L189-L191
> >> >>>
> >> >>> And untested.
> >> >>>
> >> >>> HTH,
> >> >>> Curtis
> >> >>>
> >> >>> On Thu, Jan 8, 2015 at 4:47 PM, Hadrien Mary <
> hadrien.mary at gmail.com>
> >> >>> wrote:
> >> >>>>
> >> >>>> After some investigations, I'm pretty sure I don't instanciate
> >> >>>> ImageRegion correctly. Doc and source code didn't help me... I
> would
> >> >>>> like to be able to crop a region in X and Y while keeping any other
> >> >>>> dimensions which could exist (C, Z, T).
> >> >>>>
> >> >>>> --
> >> >>>> Hadrien Mary
> >> >>>>
> >> >>>>
> >> >>>>
> >> >>>> On Thu, Jan 8, 2015 at 10:22 PM, Hadrien Mary
> >> >>>> <hadrien.mary at gmail.com>
> >> >>>> wrote:
> >> >>>> > Thank you Curtis for the answer.
> >> >>>> >
> >> >>>> > I tried to apply what you told me and the code now works without
> >> >>>> > error. However the saved cropped image is not cropped (same size
> as
> >> >>>> > original) and pixel values are modified).
> >> >>>> >
> >> >>>> > Script:
> >> >>>> >
> >> >>>> > from io.scif.config import SCIFIOConfig
> >> >>>> > from io.scif.img import ImageRegion
> >> >>>> > from io.scif.img import ImgOpener
> >> >>>> > from io.scif.img import ImgSaver
> >> >>>> >
> >> >>>> > fname = "/home/hadim/original.ome.tif"
> >> >>>> > target = "/home/hadim/cropped.ome.tif"
> >> >>>> >
> >> >>>> > config = SCIFIOConfig()
> >> >>>> > region = ImageRegion(dict(x=2, y=2, width=10, height=10))
> >> >>>> > config.imgOpenerSetRegion(region)
> >> >>>> >
> >> >>>> > opener = ImgOpener()
> >> >>>> > imps = opener.openImgs(fname, config)
> >> >>>> > imp = imps[0]
> >> >>>> >
> >> >>>> > print(imps)
> >> >>>> >
> >> >>>> > saver = ImgSaver()
> >> >>>> > saver.saveImg(target, imp)
> >> >>>> >
> >> >>>> > Thanks again for your time. Don't be sorry if you don't have time
> >> >>>> > to
> >> >>>> > write an example.
> >> >>>> >
> >> >>>> > I will be happy to provide some python/scifio examples scripts.
> >> >>>> >
> >> >>>> >
> >> >>>> > --
> >> >>>> > Hadrien Mary
> >> >>>> >
> >> >>>> > Ph.D student in Biology
> >> >>>> > Tournier-Gachet Team
> >> >>>> > CNRS - LBCMCP - UMR 5088
> >> >>>> >
> >> >>>> > Université de Toulouse - Bât. 4R3B1
> >> >>>> > 118, route de Narbonne - 31062 Toulouse
> >> >>>> >
> >> >>>> >
> >> >>>> > On Thu, Jan 8, 2015 at 9:56 PM, Curtis Rueden <ctrueden at wisc.edu
> >
> >> >>>> > wrote:
> >> >>>> >> Hi Hadrien,
> >> >>>> >>
> >> >>>> >>> I tried to use SCIFIO to write cropped image on disk but it
> doesn
> >> >>>> >>> not
> >> >>>> >>> work.
> >> >>>> >>
> >> >>>> >> The error you see is because SCIFIO operates on ImgLib2 data
> >> >>>> >> structures, not
> >> >>>> >> ImagePlus objects.
> >> >>>> >>
> >> >>>> >>> Is there is any “easy” alternative to BF setCropRegion function
> >> >>>> >>> in
> >> >>>> >>> SCIFIO ?
> >> >>>> >>
> >> >>>> >> Yes: you create a SCIFIOConfig, calling imgOpenerSetRegion [1]
> on
> >> >>>> >> it,
> >> >>>> >> then
> >> >>>> >> pass it as an argument to the ImgOpener. You'll get back an
> >> >>>> >> ImgLib2
> >> >>>> >> data
> >> >>>> >> object which can then be fed to the SCIFIO ImgSaver.
> >> >>>> >>
> >> >>>> >> I'm sorry that I don't have time to whip up an example for you
> >> >>>> >> right
> >> >>>> >> now. It
> >> >>>> >> would be great to add more SCIFIO tutorials [2] that use the
> >> >>>> >> ImgOpener
> >> >>>> >> and
> >> >>>> >> ImgSaver, since they are much higher level APIs akin to the
> >> >>>> >> Bio-Formats
> >> >>>> >> "BF"
> >> >>>> >> functionality... please feel welcome to contribute some!
> >> >>>> >>
> >> >>>> >> Regards,
> >> >>>> >> Curtis
> >> >>>> >>
> >> >>>> >> [1]
> >> >>>> >>
> >> >>>> >>
> >> >>>> >>
> http://javadoc.imagej.net/SCIFIO/io/scif/config/SCIFIOConfig.html#imgOpenerSetRegion(io.scif.img.ImageRegion)
> >> >>>> >> [2] https://github.com/scifio/scifio-tutorials
> >> >>>> >>
> >> >>>> >> On Thu, Jan 8, 2015 at 2:18 PM, Hadrien Mary
> >> >>>> >> <hadrien.mary at gmail.com>
> >> >>>> >> wrote:
> >> >>>> >>>
> >> >>>> >>> Hi,
> >> >>>> >>>
> >> >>>> >>> I am writing a python macro which iterate over all rois in ROI
> >> >>>> >>> Manager
> >> >>>> >>> and then use setCropRegion function from bioformat plugin to
> open
> >> >>>> >>> a
> >> >>>> >>> cropped region of an image.
> >> >>>> >>>
> >> >>>> >>> I tried to use SCIFIO to write cropped image on disk but it
> doesn
> >> >>>> >>> not
> >> >>>> >>> work.
> >> >>>> >>>
> >> >>>> >>> (I am using an updated version of Fiji.)
> >> >>>> >>>
> >> >>>> >>> Here is my script:
> >> >>>> >>>
> >> >>>> >>> from ij.plugin.frame import RoiManager
> >> >>>> >>> from ij import IJ
> >> >>>> >>> from io.scif.img import ImgSaver
> >> >>>> >>>
> >> >>>> >>> from loci.plugins import BF
> >> >>>> >>> from loci.plugins.in import ImporterOptions
> >> >>>> >>> from loci.common import Region
> >> >>>> >>>
> >> >>>> >>> import os
> >> >>>> >>>
> >> >>>> >>> # Get current image filename
> >> >>>> >>> imp = IJ.getImage()
> >> >>>> >>> f = imp.getOriginalFileInfo()
> >> >>>> >>> fname = os.path.join(f.directory, f.fileName)
> >> >>>> >>>
> >> >>>> >>> IJ.log('Image filename is %s' % fname)
> >> >>>> >>>
> >> >>>> >>> # Iterate over all ROIs from ROI Manager
> >> >>>> >>> rois = RoiManager.getInstance().getRoisAsArray()
> >> >>>> >>> for i, roi in enumerate(rois):
> >> >>>> >>>
> >> >>>> >>>     crop_id = i +1
> >> >>>> >>>     IJ.log("Opening crop %i / %i" % (crop_id, len(rois)))
> >> >>>> >>>
> >> >>>> >>>     bounds = roi.getBounds()
> >> >>>> >>>
> >> >>>> >>>     x = bounds.x
> >> >>>> >>>     y = bounds.y
> >> >>>> >>>     w = bounds.width
> >> >>>> >>>     h = bounds.height
> >> >>>> >>>
> >> >>>> >>>     # Import only cropped region of the image
> >> >>>> >>>     options = ImporterOptions()
> >> >>>> >>>     options.setCrop(True)
> >> >>>> >>>     options.setCropRegion(0, Region(x, y, w, h))
> >> >>>> >>>     options.setId(fname)
> >> >>>> >>>     imps = BF.openImagePlus(options)
> >> >>>> >>>
> >> >>>> >>>     imp = imps[0]
> >> >>>> >>>     imp.show()
> >> >>>> >>>
> >> >>>> >>>     crop_basename = "crop%i_%s" % (crop_id, f.fileName)
> >> >>>> >>>     crop_fname = os.path.join(f.directory, crop_basename)
> >> >>>> >>>     imp.setTitle(crop_basename)
> >> >>>> >>>
> >> >>>> >>>     # Save image
> >> >>>> >>>     IJ.log("Saving crop to %s" % crop_fname)
> >> >>>> >>>     saver = ImgSaver()
> >> >>>> >>>     saver.saveImg(crop_basename, imp)
> >> >>>> >>>
> >> >>>> >>> IJ.log('Done')
> >> >>>> >>>
> >> >>>> >>> It fails with this error:
> >> >>>> >>>
> >> >>>> >>> Traceback (most recent call last):
> >> >>>> >>>   File "/home/hadim/local/Fiji.app/plugins/Crop_Multi_Roi.py",
> >> >>>> >>> line
> >> >>>> >>> 49, in <module>
> >> >>>> >>>     saver.saveImg(crop_basename, imp)
> >> >>>> >>> TypeError: saveImg(): 1st arg can't be coerced to
> io.scif.Writer,
> >> >>>> >>> String
> >> >>>> >>>
> >> >>>> >>> at org.python.core.Py.TypeError(Py.java:235)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.python.core.PyReflectedFunction.throwError(PyReflectedFunction.java:209)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.python.core.PyReflectedFunction.throwBadArgError(PyReflectedFunction.java:312)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.python.core.PyReflectedFunction.throwError(PyReflectedFunction.java:321)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.python.core.PyReflectedFunction.__call__(PyReflectedFunction.java:167)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.python.core.PyReflectedFunction.__call__(PyReflectedFunction.java:204)
> >> >>>> >>> at org.python.core.PyObject.__call__(PyObject.java:422)
> >> >>>> >>> at org.python.core.PyObject.__call__(PyObject.java:426)
> >> >>>> >>> at org.python.core.PyMethod.__call__(PyMethod.java:139)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.python.pycode._pyx7.f$0(/home/hadim/local/Fiji.app/plugins/Crop_Multi_Roi.py:51)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.python.pycode._pyx7.call_function(/home/hadim/local/Fiji.app/plugins/Crop_Multi_Roi.py)
> >> >>>> >>> at org.python.core.PyTableCode.call(PyTableCode.java:165)
> >> >>>> >>> at org.python.core.PyCode.call(PyCode.java:18)
> >> >>>> >>> at org.python.core.Py.runCode(Py.java:1275)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.scijava.plugins.scripting.jython.JythonScriptEngine.eval(JythonScriptEngine.java:76)
> >> >>>> >>> at org.scijava.script.ScriptModule.run(ScriptModule.java:175)
> >> >>>> >>> at org.scijava.module.ModuleRunner.run(ModuleRunner.java:167)
> >> >>>> >>> at org.scijava.module.ModuleRunner.call(ModuleRunner.java:126)
> >> >>>> >>> at org.scijava.module.ModuleRunner.call(ModuleRunner.java:65)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> org.scijava.thread.DefaultThreadService$2.call(DefaultThreadService.java:164)
> >> >>>> >>> at java.util.concurrent.FutureTask.run(FutureTask.java:262)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
> >> >>>> >>> at
> >> >>>> >>>
> >> >>>> >>>
> >> >>>> >>>
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
> >> >>>> >>> at java.lang.Thread.run(Thread.java:745)
> >> >>>> >>>
> >> >>>> >>> Three questions:
> >> >>>> >>>
> >> >>>> >>> how can I write cropped image (I need to write them as OME Tiff
> >> >>>> >>> so I
> >> >>>> >>> need BF or SCIFIO).
> >> >>>> >>> Is there is any “easy” alternative to BF setCropRegion function
> >> >>>> >>> in
> >> >>>> >>> SCIFIO
> >> >>>> >>> ?
> >> >>>> >>> I am currently using RoiManager.getInstance().getRoisAsArray()
> to
> >> >>>> >>> retrieve all rois. How can I directly get a ROIManager instance
> >> >>>> >>> from
> >> >>>> >>> RoiSet.zip file ?
> >> >>>> >>>
> >> >>>> >>> Any help would be very appreciated.
> >> >>>> >>>
> >> >>>> >>> Thanks !
> >> >>>> >>>
> >> >>>> >>> —
> >> >>>> >>> Hadrien Mary
> >> >>>> >>>
> >> >>>> >>> _______________________________________________
> >> >>>> >>> ImageJ-devel mailing list
> >> >>>> >>> ImageJ-devel at imagej.net
> >> >>>> >>> http://imagej.net/mailman/listinfo/imagej-devel
> >> >>>> >>
> >> >>>> >>
> >> >>>
> >> >>>
> >>
> >> _______________________________________________
> >> ImageJ-devel mailing list
> >> ImageJ-devel at imagej.net
> >> http://imagej.net/mailman/listinfo/imagej-devel
> >
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://imagej.net/pipermail/imagej-devel/attachments/20150113/2108b222/attachment-0001.html>


More information about the ImageJ-devel mailing list