[scikit-image] 转发：回复：Re: local maxima improvements

[Yann GAVET]

Hello all,

actually, local_maxima should be equivalent to h-maxima, with h=1

h-maxima (and also local_maxima) give non ponctual locations. There 
might be a need for a function of ultimate erosion. Does it exist 
somewhere ?

I totally agree with you, the official watershed demo should be modified.

Yann GAVET
Assistant Professor - Ecole des Mines de Saint-Etienne
158 Cours Fauriel, 42023 Saint-Etienne cedex 2, FRANCE
Tel: (33) - 4 7742 0170


On 04/12/2018 09:08 AM, imagepy at sina.com wrote:
> Hi，everyone
> 
> I think we should not use peak_local_max for find watershed's seeds. why 
> not use h_maxima? which can give a h tolerance.
> I think if we should replace it in the official demo? It would cause a 
> misunderstanding.
> 
> And scikit-image's h_maxima, h_minima is very slow. here I implements 
> one with numba, 
> https://github.com/Image-Py/imagepy/blob/master/imagepy/ipyalg/hydrology/findmax.py. 
> you can see if it is useful.
> 
> yxdragon
> ----- 原始邮件 -----
> 发件人：Stefan van der Walt <stefanv at berkeley.edu>
> 收件人："Mailing list for scikit-image (http://scikit-image.org)" 
> <scikit-image at python.org>
> 主题：Re: [scikit-image] local maxima improvements
> 日期：2018年04月12日 03点08分
> 
> 
> On Wed, 11 Apr 2018 12:44:45 +1000, Juan Nunez-Iglesias wrote:
>  > In [7]: image
>  > Out[7]:
>  > array([[ 0.,  0.,  0.,  0.,  0.,  0.],
>  >        [ 0.,  1.,  0.,  0.,  0.,  0.],
>  >        [ 0.,  0.,  0.,  0.,  0.,  0.],
>  >        [ 2.,  2.,  2.,  4.,  4.,  2.],
>  >        [ 2.,  2.,  2.,  4.,  4.,  2.],
>  >        [ 2.,  2.,  2.,  2.,  2.,  2.]])
>  >
>  > In [15]: feature.peak_local_max(image)
>  > In [17]: image_peak[tuple(feature.peak_local_max(image).T)] = 1
>  >
>  > In [18]: image_peak
>  > Out[18]:
>  > array([[ 0.,  0.,  0.,  0.,  0.,  0.],
>  >        [ 0.,  1.,  0.,  0.,  0.,  0.],
>  >        [ 0.,  0.,  0.,  0.,  0.,  0.],
>  >        [ 0.,  1.,  0.,  1.,  1.,  0.],
>  >        [ 0.,  1.,  0.,  1.,  1.,  0.],
>  >        [ 0.,  0.,  0.,  0.,  0.,  0.]])
> That output in column 1 looks highly suspect! This is a great example
> for a regression test, thanks Yann.
> Stéfan
> _______________________________________________
> scikit-image mailing list
> scikit-image at python.org
> https://mail.python.org/mailman/listinfo/scikit-image
> 
> 
> _______________________________________________
> scikit-image mailing list
> scikit-image at python.org
> https://mail.python.org/mailman/listinfo/scikit-image
>