ValueError when using flood_fill on Standard RGB Image about scikit-image HOT 2 CLOSED

Hey @tokiAi, I think thats because flood_fill doesn't support images with a color channel directly und thus it expects a third coordinate for the provided seed that corresponds to the color channel dimension. Instead, use flood to create a mask first:

import skimage as ski
import matplotlib.pyplot as plt

image = ski.data.astronaut()
image_gray = ski.color.rgb2gray(image)
seed_point = (100, 100)
new_color = (0, 255, 0)

mask = ski.segmentation.flood(image_gray, seed_point, tolerance=30 / 256)
image[mask, :] = new_color  # Last dimension is color channel

fig, ax = plt.subplots()
ax.imshow(mask)
fig.show()

Note, that I rgb2gray scales the value range from (256, 0] to [1., 0.]. So I adjusted the tolerance accordingly.

Does that help you?

from scikit-image.

Thank you!

from scikit-image.