Shape analysis

[Eduardo Ortega Delgado]

Hello to everybody
I am new with ImageJ list.
I am interested to use ImageJ for shape analysis.
I already have Fourier plugin, but do not know how to use it.
I am interested to compare shapes of botanical organs, lets say leaves,
flowers, etc.
I will appreciate to receive any information on how to use ImageJ for that
purpose of recomend me any web page were I can learn or a tutorial for
that purpose.
Best greetings and Thanks in advances
Eduardo Ortega

[David Randell]

On Sunday 08 June 2008 02:38:25 Eduardo Ortega Delgado wrote:

> Hello to everybody
> I am new with ImageJ list.
> I am interested to use ImageJ for shape analysis.
> I already have Fourier plugin, but do not know how to use it.
> I am interested to compare shapes of botanical organs, lets say leaves,
> flowers, etc.
> I will appreciate to receive any information on how to use ImageJ for that
> purpose of recomend me any web page were I can learn or a tutorial for
> that purpose.
> Best greetings and Thanks in advances
> Eduardo Ortega

Hi Eduardo,

Search for "Shock Graphs" and "medial-axis transform" using Google for a
starter. Both typically take a binary (silhouette) image or simple
non-intersecting closed curve and reduces the profile of a 2D object to a
skeleton, the topology/morphology of which can then be used to classify 2D
shapes.

For a theoretical paper on this, see:
http://www.cs.toronto.edu/~sven/Papers/iccv98-shocks.pdf

If you look in ImageJ you will find a mathematical morphology routine that
will extract a skeleton from a silhouette. But its worth noting different
skeletons can be extracted from the same image, depending on the algorithm
used.  In this respect look up the plugins: BinaryThin_.class and  
BinaryThin2_.class in Gabriel Landini's software written for ImageJ:

http://www.dentistry.bham.ac.uk/landinig/software/software.html

I know one person who successfully used shock-graphs to classify leaf
silhouettes, but do not have a reference to hand. As this was for part of a
thesis project, it may not have been published.

Best regards,
Dave Randell

[Eduardo Ortega Delgado]

Thank you David!
I will follow your advices, which I appreciate very much.
Best greetings,
Eduardo

Dr. Eduardo Ortega
Profesor
Lab Fisiología Vegetal
Universidad de La Habana
Habana4, Cuba



Mensaje citado por David Randell <[hidden email]>:

> On Sunday 08 June 2008 02:38:25 Eduardo Ortega Delgado wrote:
> > Hello to everybody
> > I am new with ImageJ list.
> > I am interested to use ImageJ for shape analysis.
> > I already have Fourier plugin, but do not know how to use it.
> > I am interested to compare shapes of botanical organs, lets say leaves,
> > flowers, etc.
> > I will appreciate to receive any information on how to use ImageJ for that
> > purpose of recomend me any web page were I can learn or a tutorial for
> > that purpose.
> > Best greetings and Thanks in advances
> > Eduardo Ortega
>
> Hi Eduardo,
>
> Search for "Shock Graphs" and "medial-axis transform" using Google for a
> starter. Both typically take a binary (silhouette) image or simple
> non-intersecting closed curve and reduces the profile of a 2D object to a
> skeleton, the topology/morphology of which can then be used to classify 2D
> shapes.
>
> For a theoretical paper on this, see:
> http://www.cs.toronto.edu/~sven/Papers/iccv98-shocks.pdf
>
> If you look in ImageJ you will find a mathematical morphology routine that
> will extract a skeleton from a silhouette. But its worth noting different
> skeletons can be extracted from the same image, depending on the algorithm
> used.  In this respect look up the plugins: BinaryThin_.class and  
> BinaryThin2_.class in Gabriel Landini's software written for ImageJ:
>
> http://www.dentistry.bham.ac.uk/landinig/software/software.html
>
> I know one person who successfully used shock-graphs to classify leaf
> silhouettes, but do not have a reference to hand. As this was for part of a
> thesis project, it may not have been published.
>
> Best regards,
> Dave Randell
>

[Thomas Boudier]

Hi,

For Fourier Shape Analysis, I never used it with real images ;-) , maybe
you can have a look at the first 3-5 values, and compare them between
leafs.

Thomas


David Randell a écrit :

> On Sunday 08 June 2008 02:38:25 Eduardo Ortega Delgado wrote:
>> Hello to everybody
>> I am new with ImageJ list.
>> I am interested to use ImageJ for shape analysis.
>> I already have Fourier plugin, but do not know how to use it.
>> I am interested to compare shapes of botanical organs, lets say leaves,
>> flowers, etc.
>> I will appreciate to receive any information on how to use ImageJ for that
>> purpose of recomend me any web page were I can learn or a tutorial for
>> that purpose.
>> Best greetings and Thanks in advances
>> Eduardo Ortega
>
> Hi Eduardo,
>
> Search for "Shock Graphs" and "medial-axis transform" using Google for a
> starter. Both typically take a binary (silhouette) image or simple
> non-intersecting closed curve and reduces the profile of a 2D object to a
> skeleton, the topology/morphology of which can then be used to classify 2D
> shapes.
>
> For a theoretical paper on this, see:
> http://www.cs.toronto.edu/~sven/Papers/iccv98-shocks.pdf
>
> If you look in ImageJ you will find a mathematical morphology routine that
> will extract a skeleton from a silhouette. But its worth noting different
> skeletons can be extracted from the same image, depending on the algorithm
> used.  In this respect look up the plugins: BinaryThin_.class and  
> BinaryThin2_.class in Gabriel Landini's software written for ImageJ:
>
> http://www.dentistry.bham.ac.uk/landinig/software/software.html
>
> I know one person who successfully used shock-graphs to classify leaf
> silhouettes, but do not have a reference to hand. As this was for part of a
> thesis project, it may not have been published.
>
> Best regards,
> Dave Randell
>
>

--
/*****************************************************/
     Thomas Boudier, MCU Université Paris 6,
     UMR 7101 / IFR 83. Bat A 328, Jussieu.
     Tel : [phone]  Fax : [phone]
/****************************************************/

[Eduardo Ortega Delgado]

Thank you. I will follow your advice.
Eduardo

Dr. Eduardo Ortega
Profesor
Lab Fisiología Vegetal
Universidad de La Habana
Habana4, Cuba



Quoting Thomas Boudier <[hidden email]>:

> Hi,
>
> For Fourier Shape Analysis, I never used it with real images ;-) , maybe
> you can have a look at the first 3-5 values, and compare them between
> leafs.
>
> Thomas
>
>
> David Randell a écrit :
> > On Sunday 08 June 2008 02:38:25 Eduardo Ortega Delgado wrote:
> >> Hello to everybody
> >> I am new with ImageJ list.
> >> I am interested to use ImageJ for shape analysis.
> >> I already have Fourier plugin, but do not know how to use it.
> >> I am interested to compare shapes of botanical organs, lets say leaves,
> >> flowers, etc.
> >> I will appreciate to receive any information on how to use ImageJ for
> that
> >> purpose of recomend me any web page were I can learn or a tutorial for
> >> that purpose.
> >> Best greetings and Thanks in advances
> >> Eduardo Ortega
> >
> > Hi Eduardo,
> >
> > Search for "Shock Graphs" and "medial-axis transform" using Google for a
> > starter. Both typically take a binary (silhouette) image or simple
> > non-intersecting closed curve and reduces the profile of a 2D object to a
> > skeleton, the topology/morphology of which can then be used to classify 2D
>
> > shapes.
> >
> > For a theoretical paper on this, see:
> > http://www.cs.toronto.edu/~sven/Papers/iccv98-shocks.pdf
> >
> > If you look in ImageJ you will find a mathematical morphology routine that
>
> > will extract a skeleton from a silhouette. But its worth noting different
> > skeletons can be extracted from the same image, depending on the algorithm
>
> > used.  In this respect look up the plugins: BinaryThin_.class and  
> > BinaryThin2_.class in Gabriel Landini's software written for ImageJ:
> >
> > http://www.dentistry.bham.ac.uk/landinig/software/software.html
> >
> > I know one person who successfully used shock-graphs to classify leaf
> > silhouettes, but do not have a reference to hand. As this was for part of a
>
> > thesis project, it may not have been published.
> >
> > Best regards,
> > Dave Randell
> >
> >
>
> --
> /*****************************************************/
>      Thomas Boudier, MCU Université Paris 6,
>      UMR 7101 / IFR 83. Bat A 328, Jussieu.
>      Tel : [phone]  Fax : [phone]
> /****************************************************/
>