Using the labels layer

In this document, you will learn about the napari Labels layer, including using the layer to display the results of image segmentation analyses, and how to manually segment images using the paintbrush and fill buckets. You will also understand how to add a labels image and edit it from the GUI and from the console.

When to use the labels layer

The labels layer allows you to take an array of integers and display each integer as a different random color, with the background color 0 rendered as transparent.

The Labels layer is therefore especially useful for segmentation tasks where each pixel is assigned to a different class, as occurs in semantic segmentation, or where pixels corresponding to different objects all get assigned the same label, as occurs in instance segmentation.

A simple example

You can create a new viewer and add an labels image in one go using the napari.view_labels method, or if you already have an existing viewer, you can add a Labels image to it using viewer.add_labels. The api of both methods is the same. In these examples we’ll mainly use add_labels to overlay a Labels image onto on image.

In this example of instance segmentation, we will find and segment each of the coins in an image, assigning each one an integer label, and then overlay the results on the original image as follows:

import napari
from skimage import data
from skimage.filters import threshold_otsu
from skimage.segmentation import clear_border
from skimage.measure import label
from skimage.morphology import closing, square, remove_small_objects

coins = data.coins()[50:-50, 50:-50]
# apply threshold
thresh = threshold_otsu(coins)
bw = closing(coins > thresh, square(4))
# remove artifacts connected to image border
cleared = remove_small_objects(clear_border(bw), 20)
# label image regions
label_image = label(cleared)

# create the viewer and add the coins image
viewer = napari.view_image(coins, name='coins')
# add the labels
labels_layer = viewer.add_labels(label_image, name='segmentation')

from napari.utils import nbscreenshot

nbscreenshot(viewer, alt_text="Segmentation of coins in an image, displayed using a labels layer")

Arguments of view_labels and add_labels

view_labels() and add_labels() accept the same layer-creation parameters.

help(napari.view_labels)

Help on function view_labels in module napari.view_layers:

view_labels(data, *, num_colors=50, features=None, properties=None, color=None, seed=0.5, name=None, metadata=None, scale=None, translate=None, rotate=None, shear=None, affine=None, opacity=0.7, blending='translucent', rendering='iso_categorical', depiction='volume', visible=True, multiscale=None, cache=True, plane=None, experimental_clipping_planes=None, title='napari', ndisplay=2, order=(), axis_labels=(), show=True) -> napari.viewer.Viewer
    Create a viewer and add a labels layer.
    
    Parameters
    ----------
    data : array or list of array
        Labels data as an array or multiscale. Must be integer type or bools.
        Please note multiscale rendering is only supported in 2D. In 3D, only
        the lowest resolution scale is displayed.
    num_colors : int
        Number of unique colors to use in colormap.
    features : dict[str, array-like] or DataFrame
        Features table where each row corresponds to a label and each column
        is a feature. The first row corresponds to the background label.
    properties : dict {str: array (N,)} or DataFrame
        Properties for each label. Each property should be an array of length
        N, where N is the number of labels, and the first property corresponds
        to background.
    color : dict of int to str or array
        Custom label to color mapping. Values must be valid color names or RGBA
        arrays.
    seed : float
        Seed for colormap random generator.
    name : str
        Name of the layer.
    metadata : dict
        Layer metadata.
    scale : tuple of float
        Scale factors for the layer.
    translate : tuple of float
        Translation values for the layer.
    rotate : float, 3-tuple of float, or n-D array.
        If a float convert into a 2D rotation matrix using that value as an
        angle. If 3-tuple convert into a 3D rotation matrix, using a yaw,
        pitch, roll convention. Otherwise assume an nD rotation. Angles are
        assumed to be in degrees. They can be converted from radians with
        np.degrees if needed.
    shear : 1-D array or n-D array
        Either a vector of upper triangular values, or an nD shear matrix with
        ones along the main diagonal.
    affine : n-D array or napari.utils.transforms.Affine
        (N+1, N+1) affine transformation matrix in homogeneous coordinates.
        The first (N, N) entries correspond to a linear transform and
        the final column is a length N translation vector and a 1 or a napari
        `Affine` transform object. Applied as an extra transform on top of the
        provided scale, rotate, and shear values.
    opacity : float
        Opacity of the layer visual, between 0.0 and 1.0.
    blending : str
        One of a list of preset blending modes that determines how RGB and
        alpha values of the layer visual get mixed. Allowed values are
        {'opaque', 'translucent', and 'additive'}.
    rendering : str
        3D Rendering mode used by vispy. Must be one {'translucent', 'iso_categorical'}.
        'translucent' renders without lighting. 'iso_categorical' uses isosurface
        rendering to calculate lighting effects on labeled surfaces.
        The default value is 'iso_categorical'.
    depiction : str
        3D Depiction mode. Must be one of {'volume', 'plane'}.
        The default value is 'volume'.
    visible : bool
        Whether the layer visual is currently being displayed.
    multiscale : bool
        Whether the data is a multiscale image or not. Multiscale data is
        represented by a list of array like image data. If not specified by
        the user and if the data is a list of arrays that decrease in shape
        then it will be taken to be multiscale. The first image in the list
        should be the largest. Please note multiscale rendering is only
        supported in 2D. In 3D, only the lowest resolution scale is
        displayed.
    cache : bool
        Whether slices of out-of-core datasets should be cached upon retrieval.
        Currently, this only applies to dask arrays.
    plane : dict or SlicingPlane
        Properties defining plane rendering in 3D. Properties are defined in
        data coordinates. Valid dictionary keys are
        {'position', 'normal', 'thickness', and 'enabled'}.
    experimental_clipping_planes : list of dicts, list of ClippingPlane, or ClippingPlaneList
        Each dict defines a clipping plane in 3D in data coordinates.
        Valid dictionary keys are {'position', 'normal', and 'enabled'}.
        Values on the negative side of the normal are discarded if the plane is enabled.
        title : string, optional
        The title of the viewer window. by default 'napari'.
    ndisplay : {2, 3}, optional
        Number of displayed dimensions. by default 2.
    order : tuple of int, optional
        Order in which dimensions are displayed where the last two or last
        three dimensions correspond to row x column or plane x row x column if
        ndisplay is 2 or 3. by default None
    axis_labels : list of str, optional
        Dimension names. by default they are labeled with sequential numbers
    show : bool, optional
        Whether to show the viewer after instantiation. by default True.
    
    
    Returns
    -------
    viewer : :class:`napari.Viewer`
        The newly-created viewer.

Labels data

The labels layer is a subclass of the Image layer and as such can support the same numpy-like arrays, including dask arrays, xarrays, and zarr arrays. A Labels layer though must be integer valued, and the background label must be 0.

Because the labels layer subclasses the image layer it inherits the great properties of the image layer, like supporting lazy loading and multiscale images for big data layers. For more information about both these concepts see the details in the image layer guide.

Creating a new labels layer

As you can edit a Labels layer using the paintbrush and fill bucket, it is possible to create a brand-new empty labels layers by clicking the new labels layer button above the layers list. The shape of the new labels layer will match the size of any currently existing image layers, allowing you to paint on top of them.

Non-editable mode

If you want to disable editing of the labels layer you can set the editable property of the layer to False.

As note in the section on 3D rendering, when using 3D rendering the labels layer is not editable. Similarly, for now, a labels layer where the data is represented as a multiscale image is not editable.

3D rendering of labels

All our layers can be rendered in both 2D and 3D mode, and one of our viewer buttons can toggle between each mode. The number of dimensions sliders will be 2 or 3 less than the total number of dimensions of the layer, allowing you to browse volumetric timeseries data and other high dimensional data.

from scipy import ndimage as ndi

blobs = data.binary_blobs(length=128, volume_fraction=0.1, n_dim=3)
viewer = napari.view_image(blobs.astype(float), name='blobs')
labeled = ndi.label(blobs)[0]
labels_layer = viewer.add_labels(labeled, name='blob ID')
viewer.dims.ndisplay = 3

# programmatically adjust the camera angle
viewer.camera.zoom = 2
viewer.camera.angles = (3, 38, 53)
nbscreenshot(viewer, alt_text="A 3D view of a labels layer on top of 3D blobs")