MathGroup Archive 2002

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Inquery(Partial Differentiation)

  • To: mathgroup at smc.vnet.net
  • Subject: [mg34062] Re: Inquery(Partial Differentiation)
  • From: Jens-Peer Kuska <kuska at informatik.uni-leipzig.de>
  • Date: Wed, 1 May 2002 08:00:33 -0400 (EDT)
  • Organization: Universitaet Leipzig
  • References: <aag9hs$1sb$1@smc.vnet.net>
  • Reply-to: kuska at informatik.uni-leipzig.de
  • Sender: owner-wri-mathgroup at wolfram.com

Hi,

load a picture you like

ll = Import["/export/home/kuska/DigitalImageProcess/tiger.tiff"]

and use a smoothed Sobel filter with

bm = ll[[1, 1]];
sobel = {{1, 0, -1},
         {2, 0, -2},
         {1, 0, -1}}/8;
gradx = ListConvolve[sobel, bm, {2, 2}];
grady = ListConvolve[Transpose[sobel], bm, {2, 2}];

and here are your derivatives in x and y direction.

ListDensityPlot[gradx, Mesh -> False];
ListDensityPlot[grady, Mesh -> False];

Regards
  Jens

Gagan Anand wrote:
> 
>     I am working to Gray-scale image files(TIFF) , I
> need to calculate the Orientiation of the Pixels , for
> that I need to calculate Directional Filed Estimation.
> For Directional Field Estimation I need to fined the
> Gradient.
> Gradient can be calculated by Partial Differentiation.
> Can you help me out to calculate Partial
> Diffrenetiation at each pixel, I am clear with the
> formulae of Partial Differentiation, but how to
> Implement it.
>              Thank You,
>                               GAGAN ANAND
> 
> __________________________________________________
> Do You Yahoo!?
> Yahoo! Health - your guide to health and wellness
> http://health.yahoo.com


  • Prev by Date: Weights syntax in NonlinearRegress/Fit
  • Next by Date: [Fwd: Mathlink formulas return different results]
  • Previous by thread: Re: Weights syntax in NonlinearRegress/Fit
  • Next by thread: [Fwd: Mathlink formulas return different results]