3

I am getting some kind of noise or unintended output from skimage.segmentation.watershed(). I am new to image processing so I don't know about other parameters of watershed().

After that when I use mark_boundaries() the same output is reflected over there too.

It would be helpful if anyone knows how to remove that dots from segmented image.

here are my code and outputs

  1. mask

    plt.imshow(binarye, cmap="gray")

enter image description here

  1. The image on which I want perform segmentation enter image description here

  2. output image

    segments_watershed = watershed(img_one, 1000, mask=binarye) plt.imshow(segments_watershed, cmap="gray") enter image description here

Varshil Patel
  • 161
  • 1
  • 7
  • That looks like a bug in how the mask and seeds interact. Could you please raise an issue on github.com/scikit-image/scikit-image/issues ? Please include your example image in raw form and the code required to get to the watershed result? Thank you! – Juan Mar 16 '19 at 00:21

1 Answers1

3

It is fixed.

Gitgub issue page

Github solution

Varshil Patel
  • 161
  • 1
  • 7