NOTICE! This is a static HTML version of a legacy ImageJ Trac ticket.

The ImageJ project now uses GitHub Issues for issue tracking.

Please file all new issues there.

Ticket #540 (closed defect: fixed)

Opened 2011-05-23T16:39:18-05:00

Last modified 2011-06-07T13:23:52-05:00

Convolution oriented plugins wrong in 8-bit case

Reported by: bdezonia Owned by: bdezonia
Priority: major Milestone: biweekly-2011: Jun-06 to Jun-17
Component: Core Version:
Severity: serious Keywords:
Cc: Blocked By:
Blocking:

Description

IJ1 has special case code for convolve3x3() on a ByteProcessor. IJ2 code was adapted from ShortProcessor and FloatProcessor. As it stands 8-bit convolutions are not the same between IJ1 & IJ2. As an example run Sharpen on a sample image such as Bridge in the two programs and see the differences.

Change History

comment:1 Changed 2011-06-07T13:23:52-05:00 by bdezonia

  • Status changed from new to closed
  • Resolution set to fixed

Further testing showed it was broken for 8 and 16 bit data. Only correct for float. Fixed convolution oriented plugins and find edges to work like IJ1. Fixed in e728eefe2ebb11a9530bf2a8b8d44ce8bb27d091.