"Snakes"

"Snakes"

Post by Dusan Pi » Wed, 06 Jul 1994 01:20:26



Can you tell me something more about algorithm called "SNAKES: Active
contour models" by Michael Kass, Andrew Witkin and Demetri Teryopoulos.
I think, that the authors are members of "Schlumberger Palo Alto
Research" in Palo Alto, California.

Thanks, Dussan Pirc

 
 
 

"Snakes"

Post by John Nag » Wed, 06 Jul 1994 11:39:53



>Can you tell me something more about algorithm called "SNAKES: Active
>contour models" by Michael Kass, Andrew Witkin and Demetri Teryopoulos.
>I think, that the authors are members of "Schlumberger Palo Alto
>Research" in Palo Alto, California.

      Kass is at Apple.  Schlumberger Palo Alto Research folded years ago.

                                        John Nagle

 
 
 

"Snakes"

Post by Kellogg Boo » Wed, 06 Jul 1994 12:08:55


|>Can you tell me something more about algorithm called "SNAKES: Active
|>contour models" by Michael Kass, Andrew Witkin and Demetri Teryopoulos.
|>I think, that the authors are members of "Schlumberger Palo Alto
|>Research" in Palo Alto, California.

Witkin is at CMU, Terzopoulos at University of Toronto.
--

 
 
 

"Snakes"

Post by Manuel Gami » Sat, 09 Jul 1994 21:05:16



Quote:(Dusan Pirc) writes:
>Can you tell me something more about algorithm called "SNAKES: Active
>contour models" by Michael Kass, Andrew Witkin and Demetri Teryopoulos.
>I think, that the authors are members of "Schlumberger Palo Alto
>Research" in Palo Alto, California.

>Thanks, Dussan Pirc

Basically Snakes uses an energy minimization principle to lock a moving
spline to the contours of an image. The energy is associated with the
gradient of the image and so exibits local maxima in the contour points.
The snake is described by a set of points which are displaced acording to
a steepest descent algorithm. Internal energies to the snake are also
calculated to give elastic properties.

In the whole it works really neat. It can become unstable when there are
very steep gradients corresponding to rapidly changing intensity levels in
the image. One has to decrease the time step for the update of
the snake position or better yet low pass the whole image to remove all that
high frequency content.  I can tell you more about it if you want since I had
to code the whole thing for a semester course

--
I've seen things you people wouldn't believe
I've watched starships on fire off the shores of Orion
I've watched C beams glitter in the dark near Tanhauser Gate
All those moments lost in time like tears in rain

 
 
 

"Snakes"

Post by Raid Al-Tah » Mon, 11 Jul 1994 09:27:22


Dear Manuel Gamito;


Quote:> I can tell you more about it if you want since I
> had to code the whole thing for a semester course

    I am interested in the subject of Snakes, and I would appreciated if
    you, or anyone else, can point to few basic references to start with.
    Do you know a site where someone can FTP "Snakes" (if it is available)?
    Is your program is dependable? Are you willing to share it with others?

    Thank you ...

          ___________________________________________________
  _______|                  Raid A. Al-Tahir                 |_______
  \      |    Department of Geodetic Science & Surveying     |          /
   \     |             The Ohio State University             |     /
    \    |                Columbus, OH 43210                 |    /

   /     |___________________________________________________|     \
  /__________)                                           (__________\

 
 
 

"Snakes"

Post by LIK M » Mon, 11 Jul 1994 12:38:46


I would also appreciate some pointers/help with coding a SNAKE routine,
especially solving for the "pentadiagonal" matrix referred to in the
original paper (A) -- how do we get the matrix coeffs?

        Much thanks in advanced.

                                                        Lik

 
 
 

"Snakes"

Post by Manuel Gami » Wed, 13 Jul 1994 17:26:20



Quote:

>    I am interested in the subject of Snakes, and I would appreciated if
>    you, or anyone else, can point to few basic references to start with.
>    Do you know a site where someone can FTP "Snakes" (if it is available)?
>    Is your program is dependable? Are you willing to share it with others?

>    Thank you ...

Well, the references I have are:

Michael Kass, Andrew Witkin, Demetri Terzopoulos, "Snakes: Active Contour
Models", Int. J. Computer Vision, 1, 1987, pp 321-331.
(The original paper)

Laurent D. Cohen, "On Active Contour Models and Ballons", CVGIP: Image
Understanding, Vol. 53, No. 2, March 1991, pp 211-218.
(Introduces extensions and improvements on the original model)

Amir Amini, Terry Weymouth, Ramesh Jain, "Using Dynamic Programming for
Solving Variational Problems in Vision", IEEE Trans. on Pattern Analysis
and Machine Intelligence, Vol. 12, No. 9, September 1990, pp 855-867.
(Makes a comparison between Snakes and Dynamic Programming for Edge
Detection Problems).

Donna Williams, Mubarak Shah, ""A Fast Algorithm for Active Contours and
Curvature Estimation", CVGIP: Image Understanding, Vol. 55, No. 1,
January 1992, pp 14-26.
(More improvements on the Snake algorithm)

The implementation I have is coded in C++ (compiled with g++ v 2.5.8, don't
know about cfront) and uses IRIS GL for graphics interaction. So my guess is
it should work properly on any SGI platform running IRIX 4.0.5F. Since
IRIX 5.0.x has moved on to OpenGL some rewrites are needed here. I never
tried it on any other platform.  I will gladly mail it to you but if more
people are interested I will make it available trough anonymous ftp.

I decided to drop the semi-implicit ODE solver used by Kass et al in their
paper since I wanted high refresh rates so a simple Euler method was
used. As a result the snake tends to oscillate a little and the edge
determination loses accuracy but I get real time interaction.

Bye

Manuel Gamito

 
 
 

"Snakes"

Post by Justin Roy Mell » Thu, 14 Jul 1994 18:51:32



|> >
|> >    I am interested in the subject of Snakes, and I would appreciated if
|> >    you, or anyone else, can point to few basic references to start with.
|> >    Do you know a site where someone can FTP "Snakes" (if it is available)?
|> >    Is your program is dependable? Are you willing to share it with others?
|> >
|> >    Thank you ...
|> >
|>
|> The implementation I have is coded in C++ (compiled with g++ v 2.5.8, don't
|> know about cfront) and uses IRIS GL for graphics interaction. So my guess is
|> it should work properly on any SGI platform running IRIX 4.0.5F. Since
|> IRIX 5.0.x has moved on to OpenGL some rewrites are needed here. I never
|> tried it on any other platform.  I will gladly mail it to you but if more
|> people are interested I will make it available trough anonymous ftp.
|>
|> I decided to drop the semi-implicit ODE solver used by Kass et al in their
|> paper since I wanted high refresh rates so a simple Euler method was
|> used. As a result the snake tends to oscillate a little and the edge
|> determination loses accuracy but I get real time interaction.
|>
|> Bye
|>
|> Manuel Gamito

I'm interested in seeing this too, so could you either mail me as well. or post an ftp site.

Thanks

Justin Mellor

 
 
 

1. Newbie help with "snake" animation

please help

I'm doing a medical animation, very familar with many 3D apps, but new
to LightWave. I'm needing to have a wire-like object with a sleeve to
follow an exact path through a curving artery then the sleeve has to
retract over the wire. I can't seem to get anything to even remotely
work, bones and null objects aren't working for me ( at least the way
I'm using them), any suggestions out there?

Todd Flom

2. Mechanical Desktop v2 and many warez for cheap price!

3. "Snakes" in image processing

4. Help : problems with pm5 after using TuneUp97

5. "Snakes"

6. Bilingual Production Artist Wanted

7. need doc on "snakes"

8. Reply to Re: Bad AVIs with 3DMAX

9. ****"""""BUG IN PROGRAM"""*****

10. """"LOOKING FOR METRAEYES MODELLER AND TURBO.PXP FOR AN EXCHANGE"""""

11. Dss Satellite "HU"&"H"card programming click link

12. ***WHICH IS BEST CAPTURE"SNAPPY""AIGOTCHA"?***

13. HELP - "mli"/"mat"