octave-bug-tracker
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Octave-bug-tracker] [bug #46124] image package: bwfill should use 4-con


From: Hartmut
Subject: [Octave-bug-tracker] [bug #46124] image package: bwfill should use 4-connectivity as default
Date: Sun, 04 Oct 2015 19:03:39 +0000
User-agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:41.0) Gecko/20100101 Firefox/41.0

URL:
  <http://savannah.gnu.org/bugs/?46124>

                 Summary: image package: bwfill should use 4-connectivity as
default
                 Project: GNU Octave
            Submitted by: hardy
            Submitted on: So 04 Okt 2015 19:03:37 GMT
                Category: Octave Forge Package
                Severity: 3 - Normal
                Priority: 5 - Normal
              Item Group: Matlab Compatibility
                  Status: None
             Assigned to: None
         Originator Name: 
        Originator Email: 
             Open/Closed: Open
         Discussion Lock: Any
                 Release: 4.0.0
        Operating System: Any

    _______________________________________________________

Details:

This can be oberved in the current image package (version 2.4.1):

As properly described in the help text of bwfill, this function uses as
default value "8-connectivity".

The Octave function bwfill only works on binary images (as its name suggests)
and substitutes the Matlab function imfill (that also works on other types of
images).

The Matlab function imfill uses as default value "4-connectivity", this can
for example be seen here:
http://de.mathworks.com/help/images/ref/imfill.html#inputarg_conn

I thinks the Matlab default behavior is MORE USEFUL. For example if you use an
edge detector (like the canny one) on an image, what you get is usually a
8-connected contour. In order to fill it (i.e. fill the background pixels
inside this contour) you then need to fill something with a 4-connected
behavior. Therefore currently the default setting in Octave FAILS on the (most
common?) usecase of filling contours from edge extraction. This is why I like
the Matlab default value better.

Could we change this behavior, to be Matlab compatible (and better! as
described above)?






    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?46124>

_______________________________________________
  Nachricht gesendet von/durch Savannah
  http://savannah.gnu.org/




reply via email to

[Prev in Thread] Current Thread [Next in Thread]