Hi Rémi!

This is a cool problem but I'm afraid I don't have a solution for you. I'm not familiar with snakes so I can't help you with the parameters.

One thing I will say is that all methods will benefit from a ground truth dataset, where you have manually painted the result you're looking for. Once you have that you can really expand your parameter and method search dramatically because you don't have to look at each answer. Do you have such a dataset?

Good luck!


On Thu, Jul 21, 2016 at 11:26 PM, Rémi Cura <remi.cura@gmail.com> wrote:
Hello everybody.
I'm a real fan of skimage,
and use it for numerous image processing stuf.

I'm stuck in a segmentation problem: I have an old Paris map, and I try to separate the streets and the buildings.
I have rough street axis.
This looks like a watershed problem, but sme of the building footprints are not closed, so the watershed starting in streets is bleeding in the buildings.

To avoid that, I was thinking of snake, starting conservatively in the streets, then make them balloon .

I can't find the parameters so the snake balloons.
Is this possible,
would you have any advice?


You received this message because you are subscribed to the Google Groups "scikit-image" group.
To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe@googlegroups.com.
To post to this group, send email to scikit-image@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/scikit-image/fb202508-b0af-4af6-a2fc-15d7db6317a1%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.