[ome-users] question about Bio-Formats.

[Melissa Linkert]

Hi Qi,

> I follow your suggestion but still face a problem as attachment. Does it
> memory problem? Thank you.

Yes, this new error message does indicate that more memory should be
allocated.   This is expected given that the heap size is set to 128MB;
as noted in the bfGetPlane output, a minimum of 512MB is suggested.

If increasing the heap size to at least 512MB does not help, please let
us know.

Regards,
-Melissa

On Wed, Apr 01, 2015 at 04:49:23PM -0400, Qi Gong wrote:
> Hello Melissa,
> 
> Thank you for your help.
> 
> I follow your suggestion but still face a problem as attachment. Does it
> memory problem? Thank you.
> 
> Sincerely,
> Qi
> 
> On Mon, Mar 30, 2015 at 3:31 PM, Melissa Linkert <
> melissa at glencoesoftware.com> wrote:
> 
> > Hi Qi,
> >
> > > I find in the web you gave me, all the sample is begin by bfopen. Thus, I
> > > try the code. But it give me some errors as attachment. The warning is
> > > about memory, do you think the error is also because that? For memory, I
> > > find we could paste the java.opts file as attachment in matlab root
> > folder
> > > to increase it. But we are building a GUI by matlab, we want to compile
> > it
> > > and send to other people. Then other user could only install MATLAB
> > > Compiler Runtime to run it. I think they will have the same problem. Does
> > > it possible they paste the java.opts file to their MATLAB Compiler
> > Runtime
> > > folder to solve the problem.
> >
> > The error you see is a result of the image size exceeding 2GB; while we
> > suggest increasing the memory in general (as noted in the warning), this
> > won't resolve the error.  To resolve the warning, Mathworks has
> > documentation
> > on using java.opts with MCR:
> >
> >
> > http://www.mathworks.com/matlabcentral/answers/99448-how-can-i-include-a-java-opts-file-in-my-matlab-compiler-generated-application
> >
> > To resolve the error though, you will need to open the image in smaller
> > tiles.
> > This can be done by starting with bfGetReader:
> >
> >
> > http://www.openmicroscopy.org/site/support/bio-formats5.0/developers/matlab-dev.html#reading-from-an-image-file
> >
> > and then using bfGetPlane to retrieve a tile.  Using the variables from
> > above page, that would look something like:
> >
> > %%%%
> > imageWidth = reader.getSizeX();
> > imageHeight = reader.getSizeY();
> > firstTile = bfGetPlane(reader, 1, 1, 1, 1024, 1024);
> > lastTile = bfGetPlane(reader, 1, imageWidth - 1023, imageHeight - 1023,
> > 1024, 1024);
> > %%%%
> >
> > As noted in bfGetPlane.m, the 5 integer parameters are the plane number,
> > X and Y coordinates of the upper-left tile corner, and width and height
> > of the tile.  We do not currently have a good Matlab example of working
> > with large images, but the corresponding Java example is (ignoring the
> > lines referencing "writer"):
> >
> >
> > http://www.openmicroscopy.org/site/support/bio-formats5.0/developers/export.html#converting-large-images
> >
> > "reader" in this example is the same class as "reader" in the above
> > Matlab-specific example.
> >
> > Hopefully that helps, but if you encounter any other problems please let
> > us know.
> >
> > Regards,
> > -Melissa
> >
> > On Mon, Mar 30, 2015 at 03:03:14PM -0400, Qi Gong wrote:
> > > Hello Lan,
> > >
> > > I find in the web you gave me, all the sample is begin by bfopen. Thus, I
> > > try the code. But it give me some errors as attachment. The warning is
> > > about memory, do you think the error is also because that? For memory, I
> > > find we could paste the java.opts file as attachment in matlab root
> > folder
> > > to increase it. But we are building a GUI by matlab, we want to compile
> > it
> > > and send to other people. Then other user could only install MATLAB
> > > Compiler Runtime to run it. I think they will have the same problem. Does
> > > it possible they paste the java.opts file to their MATLAB Compiler
> > Runtime
> > > folder to solve the problem. Thank you.
> > >
> > > On Mon, Mar 30, 2015 at 11:31 AM, Munro, Ian <i.munro at imperial.ac.uk>
> > wrote:
> > >
> > > > My apologies for mistyping your name.
> > > >
> > > > Ian
> > > >
> > > >
> > > > _______________________________________________
> > > > ome-users mailing list
> > > > ome-users at lists.openmicroscopy.org.uk
> > > > http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-users
> > > >
> > >
> > >
> > >
> > > --
> > > mobile phone: 2024060862
> >
> >
> >
> > > _______________________________________________
> > > ome-users mailing list
> > > ome-users at lists.openmicroscopy.org.uk
> > > http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-users
> >
> >
> 
> 
> -- 
> mobile phone: 2024060862