#305 polyfill/polyfillv includes boundary pixels

closed-accepted
nobody
other (94)
1
2012-10-06
2012-06-14
No

Opening this to track an issue reported on
the PDL mailing lists. Comparing the new
pnpoly routine against the existing
polyfill/polyfillv routines showed that the
pixels affected/selected are not quite
right in that some of the pixels on the
edge of the polygon are being included.

See the report and discussion at:
http://mailman.jach.hawaii.edu/pipermail/perldl/2012-June/006943.html

The +-1 pixel difference is small but if you
need exact results, use pnpoly over the
polyfill variants. NOTE: pnpoly is slower
and more memory intensive that the polyfill
routines which have less temps and are
all PP implemented.

Discussion

  • Chris Marshall

    Chris Marshall - 2012-06-14

    Test routine showing difference (by Tim Haines)

     
  • Tim

    Tim - 2012-07-03

    I submitted a patch for this.

    See pdl-porters mailing iist thread for more details (http://mailman.jach.hawaii.edu/pipermail//pdl-porters/2012-June/004786.html).

    Thanks.

    - Tim Haines

     
  • Chris Marshall

    Chris Marshall - 2012-07-03

    Could you attach the candidate patch to this ticket? Thanks, Chris.

     
  • Chris Marshall

    Chris Marshall - 2012-09-02

    Bug fixed in Git.
    Thanks for reporting the problem!

     
  • Chris Marshall

    Chris Marshall - 2012-09-02
    • status: open --> pending
     
  • Chris Marshall

    Chris Marshall - 2012-09-02

    A PP implementation of pnpoly has been contributed by Tim Haines and is available as an option for polyfill and polyfillv. The new code should be available in the next PDL developers release.

     
  • Chris Marshall

    Chris Marshall - 2012-10-06

    Bug fixed in Git.
    Thanks for reporting the problem!

     
  • Chris Marshall

    Chris Marshall - 2012-10-06
    • status: pending --> closed-accepted
     

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks