|
From: | Hartmut |
Subject: | [Octave-bug-tracker] [bug #51884] [octave forge] (image) bwmorph(blackNwhiteMatrix, "thicken", 1) returns negated result (+ strange behavior at border) |
Date: | Tue, 4 Dec 2018 13:58:44 -0500 (EST) |
User-agent: | Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:63.0) Gecko/20100101 Firefox/63.0 |
Follow-up Comment #17, bug #51884 (project octave): The Octave implementation of bwmorph(thicken) relies on the details in the Pratt book (see literature in comment #15), as stated in our source code (__conditional_mark_patterns_lut_fun__.m). But the Matlab version states in its help page to rely on the details in the Lam book (see their section on "thin"). So I would assume they are just doing a slightly different thing. I myself can't spend the effort to fully re-implement the Octave version, using a different algorithm (as stated in the Lam book) in order to become fully compatible in this case. Nevertheless I think that the current patch (file #41737) is an improvement to the current Octave implementation. The original ideas come from Hans, a new contributor to the image package. Using this patch, our current code will better do what it claims to do in its help text. Is this enough to make use of this code? My personal opinion would be yes. _______________________________________________________ Reply to this item at: <https://savannah.gnu.org/bugs/?51884> _______________________________________________ Message sent via Savannah https://savannah.gnu.org/
[Prev in Thread] | Current Thread | [Next in Thread] |