Yeah, this works well! BTW, the unwanted thicker bounds could be removed by a simple binary dialation. Thanks! On Thu, Mar 28, 2013 at 9:14 PM, Juan Nunez-Iglesias wrote:

I think the following should work:

# ws is your watershed label mapfrom scipy import ndimage as nd bound = nd.grey_erosion(ws) != nd.grey_dilation(ws)

The only bad thing about this is that you'll get slightly thick boundaries. I'm not sure if that's a problem for you.

On Thu, Mar 28, 2013 at 10:37 AM, Zetian Yang wrote:

...

Thanks Walt,

I need a binary image in which the edges of every labeled region are set to 1. I've found the `find_boundaries` function in the skimage.segmentation module, but its outcome didn't fit my requirement.

Currently, I'm using the following algorithm to solve my question.

``` bound = np.zeros(data.shape) labels = data.max() for label in labels: label_data = data==label bound += label_data - scipy.ndimage.binary_erosion(label_data) ```

It seemed work, but I'm not sure its correctness and is there a more efficient method?

On Thu, Mar 28, 2013 at 12:00 AM, Stéfan van der Walt wrote:

...

Hi Zetian

On Wed, Mar 27, 2013 at 5:07 AM, Zetian Yang wrote:

...

I have been trying the watershed algorithm in the skimage package and it is really fantastic. Recently I have a problem where the edge of one segmented image is need. Is there a convenient way to extract edges of the watersheded result?

I'm glad you find the package useful! Do you need the edges as coordinates, or do you need a bitmap of the edges? We have marching squares for contour finding, edge detection, etc.

Stéfan

-- 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. For more options, visit https://groups.google.com/groups/opt_out.

-- 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. For more options, visit https://groups.google.com/groups/opt_out.

-- 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. For more options, visit https://groups.google.com/groups/opt_out.