Ticket #248 (closed defect: fixed)
Opened 2010-10-08T11:30:51-05:00
Last modified 2011-05-31T16:45:03-05:00
Bio-Formats ITK plug-in - populate physical pixel sizes
Reported by: | hinerm | Owned by: | hinerm |
---|---|---|---|
Priority: | major | Milestone: |
|
Component: | SCIFIO | Version: | |
Severity: | serious | Keywords: | |
Cc: | Blocked By: | ||
Blocking: |
Description
Currently this is not done due to invalid memory access errors on some systems (e.g., OS X 10.5) when such code is enabled.
Change History
comment:2 Changed 2011-05-05T14:26:32-05:00 by hinerm
- Status changed from accepted to closed
- Resolution set to fixed
Discovered that calls to getPixelPhysicalSizeX/Y/Z/TimeIncrement could come back NaN if the values weren't set in the metadata, causing crashes when doubleValue() was called.
Added checks for NaN in BFITK Jace and JNI. The physical pixel sizes are now set to 1 if determined to be NaN in the metadata.
comment:3 Changed 2011-05-23T15:01:53-05:00 by hinerm
- Status changed from closed to reopened
- Resolution fixed deleted
In the Jace implementation there is a NullPointerException at the first getPixelsPhysicalSize. This doesn't stop the image from opening but needs to be resolved.
The JNI implementation seems to be functional
Note that the Pipes implementation already had code implemented to perform this functionality.
comment:4 Changed 2011-05-27T15:47:17-05:00 by curtis
- Milestone changed from biweekly-2011: Apr-25 to May-06 to biweekly-2011: May-23 to Jun-03
Mark, could you please post the stack trace of the NPE?
comment:5 Changed 2011-05-31T16:45:03-05:00 by hinerm
- Status changed from reopened to closed
- Resolution set to fixed
Issue resolved in ome:changeset:1428d93884503b9e9b5422b9fd0d751a4c08f7f1/bioformats.git