[ImageJ-bugs] [Bug 905] New: Threshold signed 16-bit images

bugzilla at fiji.sc bugzilla at fiji.sc
Fri Aug 15 14:21:20 CDT 2014


http://fiji.sc/bugzilla/show_bug.cgi?id=905

            Bug ID: 905
           Summary: Threshold signed 16-bit images
           Product: Fiji
           Version: unspecified
          Hardware: Macintosh
                OS: Mac OS
            Status: NEW
          Severity: major
          Priority: P5
         Component: ImageJ2
          Assignee: imagej-bugs at imagej.net
          Reporter: glenmac at uw.edu

Created attachment 199
  --> http://fiji.sc/bugzilla/attachment.cgi?id=199&action=edit
Huygens Pro OME XML file

This may be more than 1 issue. 

Source files are signed 16-bit OME XML from Huygens Professional.
Opening by SCIFIO resets the histogram to 0-maximum, Bio-Formats displays the
image as -32768 to 32767 until I hit 'Reset' in B&C.

calling Threshold from menu or run("Threshold...") from macro will set by the
default method.  But, checking the box for 'Stack histogram' renders the
channel black with min = 32768, max=32767 until hit 'reset' in threshold window
or B&C window.  When the channel is black, moving to a different channel and
returning will restore the display but cancels the threshold.  

If I run setAutoThreshold("Default dark stack"); from the macro then it sets
min = 32768, max=32767 and then all slices in the channel are masked as red. 
Smetimes If I click on 'Auto', this resets with the maybe the stack histogram. 
Other times I get min= -32768 and max= -max histogram (negative).

Similar results with import by Bio-Formats, although it does not make the
initial histogram reset.  

I'd like to call setAutoThreshold because I can specify the stack histogram. 
The run(Threshold) does not allow calling the stack histogram.

I've attached a file from Huygens.
Regards,
glen

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://imagej.net/pipermail/imagej-bugs/attachments/20140815/997e4127/attachment.html>


More information about the Imagej-bugs mailing list