Glasbey colormap

A long-requested feature [1] for napari was to display labels/segmentation layers using a well-known discrete colormap such as Glasbey [2].

In this example, we demonstrate displaying segmentations using custom colormaps with the help of the glasbey Python library [3], which you can install with your favorite Python package manager, such as pip or conda. We display a segmentation using the napari built-in labels colormap, the original Glasbey colormap, a more modern version produced by limiting the lightness and chroma and optimizing for colorblind-safety, and finally with the matplotlib tab10 colormap.

[1]

napari/napari#454

[2]

Colour displays for categorical images. Chris Glasbey, Gerie van der Heijden, Vivian F. K. Toh, and Alision Gray. (2007) DOI:10.1002/col.20327

[3]

lmcinnes/glasbey

Tags: layers, visualization-basic

import glasbey
import numpy as np
from skimage import data
from skimage.filters import threshold_otsu
from skimage.measure import label
from skimage.morphology import closing, remove_small_objects
from skimage.segmentation import clear_border

import napari

image = data.coins()[50:-50, 50:-50]

First, we segment the image.

# apply threshold
thresh = threshold_otsu(image)
bw = closing(image > thresh, footprint=np.ones((4, 4), dtype=bool))

# remove artifacts connected to image border
cleared = remove_small_objects(clear_border(bw), 20)

# label image regions
label_image = label(cleared).astype("uint8")

Then, we create two color palettes using the glasbey library. One with the original glasbey parameters and 256 colors, and a more modern one with better lightness and chroma bounds for a less glary look.

# original glasbey
glas = glasbey.create_palette(256)

# more optimized glasbey
glas19mid = glasbey.create_palette(
        19,
        lightness_bounds=(20, 60), chroma_bounds=(40, 50),
        colorblind_safe=True,
        )

Finally, we display the coins image and the overlaid segmentation. We do this in two viewers to show both colormaps.

viewer, image_layer = napari.imshow(image, name='coins')

# add the labels
label_layer_glas = viewer.add_labels(
        label_image, name='segmentation', colormap=glas
        )

viewer2, image_layer2 = napari.imshow(image, name='coins')

label_layer_modern = viewer2.add_labels(
        label_image, name='segmentation-glasbey-19-mid-chroma', colormap=glas19mid,
        )


if __name__ == '__main__':
    napari.run()