Re: Sharpening Filter

Hi

I do not know the name of this filter but for some reason it reminds me of the the kuwahara filter. The kuwahara filter splits the structuring element into 4 overlapping elements (NW, NE, SW, SE) like this:

123 12 23
456 -> 45 56
789
45 56
78 89

and computes the variance of each of the 4 elements. The filtered value of the center pixel 5 is then the mean of the element (NW, NE, SW, SE) with the smallest variance.

Kind regards
Allan Lyckegaard

thomas_m_richardson@xxxxxxxxx wrote:
A colleague has proposed the following sharpening filter to me:

Use a 3x3 structuring element.

Compute the mean, min, and max on the structuring element.

The filtered result is

if (MEAN - MIN) > (MAX - MEAN) then
RESULT = MAX
else
RESULT = MIN

One could think of this as an adaptive erosion/dialation.

Does this filter have a common name? Is it well known? Is it similar to
any well known filters?

Thanks in advance, Tom

.

Relevant Pages

  • Sharpening Filter
    ... A colleague has proposed the following sharpening filter to me: ... Compute the mean, min, and max on the structuring element. ...
    (sci.image.processing)
  • Re: strange behaviour of ntp peerstats entries.
    ... variance truly does minimize the offset variance. ... the filter reduces the variance by some 10 ... delay variance IS reduced, ...
    (comp.protocols.time.ntp)
  • Re: Lowpass filtering for OFDM introduces interference?
    ... Evaluating this at some arbitrary k = k_0, this relationship is clearly not affected by the size of the DFT, nor by the amount of noise at any other k. ... variance of vshould be reduced leading to a reduced noise variance of ... Because each bin of the DFT is a matched correlation filter. ...
    (comp.dsp)
  • Re: Efficient Moving Average and Moving Variance Calculations
    ... you need to learn how to do "ASCII math" (a relative of "ASCII ... brackets instead of parenths to group together multiple terms, ... variance, as computed by the mean square of the samples, is a slightly ... filter *does* have a forgetting factor, and, even though i do not have ...
    (comp.dsp)
  • Re: strange behaviour of ntp peerstats entries.
    ... variance truly does minimize the offset variance. ... the filter reduces the variance by some 10 ... A 9ms delay from a collision introduces a 4ms change in the ...
    (comp.protocols.time.ntp)