Hi Yan, I actually really want to pilfer a lot of your code, but I find it hard to understand/follow. If you have a chance, annotations on your code would make it really useful to put it into scikit-image (I’d be happy to write tests etc.). I specifically am interested in your fast morphology and your ridge detection algorithms.
Some suggestions: - Increase variable names to be more descriptive. For example in findmax, what is sta? I have no idea. idx is used in various places with different meanings. etc. - For each function, add a docstring describing what it does, perhaps with some doctest example. Some functions are self explanatory but many are not. - add a few comments in the code for the critical parts of each algorithm, possibly with references back to a paper/wikipedia entry I know you are super busy but at this stage you are the best placed to do this work — it would take me a long time to reverse engineer the algorithm based on your code! Juan. On 12 Apr 2018, 5:08 PM +1000, imag...@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 <stef...@berkeley.edu> > 收件人："Mailing list for scikit-image (http://scikit-image.org)" > <scikit-image@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@python.org > https://mail.python.org/mailman/listinfo/scikit-image > _______________________________________________ > scikit-image mailing list > scikit-image@python.org > https://mail.python.org/mailman/listinfo/scikit-image
_______________________________________________ scikit-image mailing list scikit-image@python.org https://mail.python.org/mailman/listinfo/scikit-image
