Ticket #1652 (reviewing defect)
Opened 2013-02-19T14:33:48-06:00
Last modified 2013-06-04T12:29:02-05:00
Resolve inconsistencies of threshold code
| Reported by: | bdezonia | Owned by: | curtis |
|---|---|---|---|
| Priority: | major | Milestone: | imagej2-b8-analysis |
| Component: | Core | Version: | |
| Severity: | serious | Keywords: | |
| Cc: | Blocked By: | ||
| Blocking: | #1678 |
Description
Currently there is an initial implementation of thresholding capabilities in IJ2. It was designed to fit current architectures. But there is an issue at the core of the design that needs to be resolved.
As designed Overlays apply to displays. I think so in the future we can support an Overlay that applies to multiple datasets within a display. However the ThresholdOverlay code is more Dataset centric. And other code (like ConvertToMask and ThresholdService) have to apply to the active dataset of a display. And our JHotDraw code might be messy if the threshold overlays apply to multiple datasets.
We need to decide if the threshold overlay is Dataset centric or Display centric. If Display centric then enhance the threshold code to work as such. CTR please comment with your thoughts and assign back if necessary.