18 Sep
2020
18 Sep
'20
9:53 a.m.
Wow, yes this is fun! Here are two (messy) data sets: https://0x0.st/ils7.png https://0x0.st/ilsh.png Each point configuration is labeled with an initial and final score. The two scores seem to be uncorrelated. This is too bad because the ellipse score should go to 1 as the ellipse approaches a circle. Is there a way to score the initial data so that we can tell when it will turn out most like a circle in the end? (I didn't read the paper too closely but did ctrl-f "circle" and didn't find anything relevant.) --Brad On Fri, Sep 18, 2020 at 9:27 AM Lucas, Stephen K - lucassk <lucassk@jmu.edu> wrote:
For those who wish to see the fun that can happen, check out http://www.cs.cornell.edu/cv/ResearchPDF/PolygonSmoothingPaper.pdf
Steve