Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.soft-sys.math.mathematica > #14291 > unrolled thread

Searching for circular structures in an image

Started byLorenz Müller <lorenz.mueller@bluewin.ch>
First post2013-04-24 11:02 +0000
Last post2013-04-26 08:24 +0000
Articles 4 — 2 participants

Back to article view | Back to comp.soft-sys.math.mathematica


Contents

  Searching for circular structures in an image Lorenz Müller <lorenz.mueller@bluewin.ch> - 2013-04-24 11:02 +0000
    Re: Searching for circular structures in an image debguy <johnandsara2@cox.net> - 2013-04-25 06:52 +0000
      Re: Searching for circular structures in an image debguy <johnandsara2@cox.net> - 2013-04-26 08:24 +0000
      Re: Searching for circular structures in an image debguy <johnandsara2@cox.net> - 2013-04-26 08:24 +0000

#14291 — Searching for circular structures in an image

FromLorenz Müller <lorenz.mueller@bluewin.ch>
Date2013-04-24 11:02 +0000
SubjectSearching for circular structures in an image
Message-ID<kl8e4c$on3$1@smc.vnet.net>
Does somebody have a Mathematica implementation  to search an image for circular structures using circular Hough transform (function or module that generates the Hough accumulator)
Thanks for any hints

[toc] | [next] | [standalone]


#14308

Fromdebguy <johnandsara2@cox.net>
Date2013-04-25 06:52 +0000
Message-ID<klajr3$3b0$1@smc.vnet.net>
In reply to#14291
Good question i'm not sure I know.

I looked on wikipedia to decipher what your asking.

There is a recent posting just days ago involving shape recognition.
However the "answer" is also not mathematical theory as,
mathematically, many kinds of errors arise in the math is very
assuming (ie, drops many sol'n).  Yet it does get one in the ball-park
easily and quickly one can say.

What I found on wiki didn't qualify as a function (none was posted,
and it wouldn't be possible to post one.  Nor did the problem or
answer wiki describes qualify as formal mathematical theory or
anything close to being a theory.  Infact it only describes a
whimsical comparison and doesn't do justice to what kind of math might
be needed.

A "generalized" sol'n might be to find linear multiple of splines in 3
space within a given percentage, not just ellipsoids in 2 space?
There are many recognition and learning algorithms.  Which brings up
an important question: what exactly do the data and results require?

It's a popular topic and Im sure google is bound to have many results
if your looking for code and algorithms.

As far as pattern recognition i'm unsure what if any of mathematica's
new features best fit.  It seems to be rather straight forward,
statistically speaking, given the previous posting.  I'm sure
mathematica will prove useful while doing it.

[toc] | [prev] | [next] | [standalone]


#14317

Fromdebguy <johnandsara2@cox.net>
Date2013-04-26 08:24 +0000
Message-ID<klddkf$d17$1@smc.vnet.net>
In reply to#14308
I'm sorry I didn't see the pages in the new online Mathematica book in
section

Image Processing & Analysis

         ie,  FindFaces, TextRecognize

Did you try these?  Anyway my version doens't have it and I have no
images needing to be analyzed.  But I bet there are book examples
which can do the task.

[toc] | [prev] | [next] | [standalone]


#14320

Fromdebguy <johnandsara2@cox.net>
Date2013-04-26 08:24 +0000
Message-ID<klddkl$d19$1@smc.vnet.net>
In reply to#14308
also excuse me i didn't (don't) count applications of rectangular to
polar coordinates, Hough, as anything new: I consider it an iterated
application of pretty simple math.  i completely didn't realize they
named it.  i see so much code.  many of mm's newer image abilities are
old news as, say, Gimp image software, goes.  though some are very
useful.

[toc] | [prev] | [standalone]


Back to top | Article view | comp.soft-sys.math.mathematica


csiph-web