[Insight-users] itkVectorFuzzyConnectednessImageFilterTest

[Sun Hao]

Hi Luis,
    I do know how to compute the covariance matrix. In the example of itkVectorFuzzyConnectednessImageFilterTest, I can compute object_cov1 and object_cov2 -- the two convariance matrixes, but I don't know how to get homo_cov. Can you tell me that? Thanks!

Regards,

Sun Hao

----- Original Message ----- 
From: "Luis Ibanez" <luis.ibanez at kitware.com>
To: "Sun Hao" <snho at sjtu.edu.cn>
Cc: <insight-users at itk.org>
Sent: Tuesday, August 03, 2004 12:41 PM
Subject: Re: [Insight-users] itkVectorFuzzyConnectednessImageFilterTest


> 
> Hi Sun,
> 
> You compute this matrix by taking a region of "example"
> pixels that are representative of the image feature
> that you want to segment.
> 
> For example if you have CT images of the liver, then
> you select a group of "liver" pixels.
> 
> Note that this method is intended to be used with
> Vector images.  Once you have a group of pixels, you
> take their "vector values" and compute their covariance
> matrix. This is a common statistical construct that you
> will find explained in any introductory book on statistics.
> 
> In a practical real application you want to do this by
> asking the user to click on representative points of
> the image. Around those points you select regions of
> a pre-specified size, take all the pixels inside those
> regions and use their Vector pixel values in order to
> compute their covariance metric.
> 
> 
>    Regards,
> 
> 
>      Luis
> 
> 
> 
> 
> ------------------------
> Sun Hao wrote:
> 
> > Hi everybody,
> > When I read  itkVectorFuzzyConnectednessImageFilterTest.cxx in ITK 
> > examples, I am puzzled by the variable 
> > homo_cov[FEATURES_NUM][FEATURES_NUM]. Can you tell me how to compute 
> > this matix? Thank you!
> > 
> > Best regards,
> > 
> >     Sun Hao
> > 
> > 
> > ------------------------------------------------------------------------
> > 
> > _______________________________________________
> > Insight-users mailing list
> > Insight-users at itk.org
> > http://www.itk.org/mailman/listinfo/insight-users
> 
> 
> 
>