unsupervised learning

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

unsupervised learning

Aloysius Phillips
Hello,

I would like to classify 2D shapes from cross sections using unsupervised learning. Does anyone have any advice about which attributes ImageJ can describe? The more the better. It doesnt matter if they are correlated. Also, any advice on unsupervised learning in general would be appreciated. Thank you.

 
Aloysius Phillips
Division of Invertebrates
American Museum of Natural History
79th street and Central Park West
10024

Voice (212) 769-5410
FAX (212) 769-5277
Reply | Threaded
Open this post in threaded view
|

Re: unsupervised learning

Gary Chinga
Try the shape descriptor plugin at http://www.gcsca.net/IJ/ 
Shapes.html. It may be a start.

Gary
www.gcsca.net


On Jun 15, 2006, at 3:51 AM, Aloysius Phillips wrote:

> Hello,
>
> I would like to classify 2D shapes from cross sections using  
> unsupervised learning. Does anyone have any advice about which  
> attributes ImageJ can describe? The more the better. It doesnt  
> matter if they are correlated. Also, any advice on unsupervised  
> learning in general would be appreciated. Thank you.
>
>
> Aloysius Phillips
> Division of Invertebrates
> American Museum of Natural History
> 79th street and Central Park West
> 10024
>
> Voice (212) 769-5410
> FAX (212) 769-5277
>
>
Reply | Threaded
Open this post in threaded view
|

Re: unsupervised learning

audrey karperien
In reply to this post by Aloysius Phillips
FracLac finds some attributes of 2d shapes, such as the fractal dimension (for complexity) and lacunarity (for heterogeneity), as well as some other morphometrics.  Its available here:
  http://rsb.info.nih.gov/ij/plugins/frac-lac.html

Aloysius Phillips <[hidden email]> wrote:
  Hello,

I would like to classify 2D shapes from cross sections using unsupervised learning. Does anyone have any advice about which attributes ImageJ can describe? The more the better. It doesnt matter if they are correlated. Also, any advice on unsupervised learning in general would be appreciated. Thank you.


Aloysius Phillips
Division of Invertebrates
American Museum of Natural History
79th street and Central Park West
10024

Voice (212) 769-5410
FAX (212) 769-5277