Colormap#

Render a single-channel image with a color map, e.g. to visualize depth, heat or edge maps.

Functions#

kornia.color.apply_colormap(input_tensor, colormap)[source]#

Apply to a gray torch.Tensor a colormap.

Try in browser

Convention:: Accepts a rank-3 (C, H, W) or rank-4 (B, C, H, W) tensor and returns (B, 3*C, H, W), with B = 1 for rank-3 input; input channel c becomes output channels 3c to 3c + 2. Integer inputs use [0, 255]. Floating inputs use [0, 1] or [0, 255] independently for each sample and channel, based on that channel’s maximum.

Parameters:

input_tensor (Tensor) – the input torch.Tensor of image.
colormap (ColorMap) – the colormap desired to be applied to the input torch.Tensor.

Return type:

Tensor

Returns:

A tensor with the applied color map.

Example

>>> input_tensor = torch.tensor([[[0, 1, 2], [15, 25, 33], [128, 158, 188]]])
>>> colormap = ColorMap(base=ColorMapType.autumn)
>>> apply_colormap(input_tensor, colormap)
tensor([[[[1.0000, 1.0000, 1.0000],
          [1.0000, 1.0000, 1.0000],
          [1.0000, 1.0000, 1.0000]],

         [[0.0000, 0.0159, 0.0159],
          [0.0635, 0.1111, 0.1429],
          [0.5079, 0.6190, 0.7302]],

         [[0.0000, 0.0000, 0.0000],
          [0.0000, 0.0000, 0.0000],
          [0.0000, 0.0000, 0.0000]]]])

Modules#

class kornia.color.ApplyColorMap(colormap)[source]#

Class for applying a colormap to images.

Try in browser

See the Convention block on apply_colormap().

Parameters:: colormap (ColorMap) – The ColorMap object to apply.
Returns:: A tensor with the applied color map.

Example

>>> input_tensor = torch.tensor([[[0, 1, 2], [15, 25, 33], [128, 158, 188]]])
>>> colormap = ColorMap(base=ColorMapType.autumn)
>>> ApplyColorMap(colormap=colormap)(input_tensor)
tensor([[[[1.0000, 1.0000, 1.0000],
          [1.0000, 1.0000, 1.0000],
          [1.0000, 1.0000, 1.0000]],

         [[0.0000, 0.0159, 0.0159],
          [0.0635, 0.1111, 0.1429],
          [0.5079, 0.6190, 0.7302]],

         [[0.0000, 0.0000, 0.0000],
          [0.0000, 0.0000, 0.0000],
          [0.0000, 0.0000, 0.0000]]]])

class kornia.color.ColorMap(base, num_colors=64, device=None, dtype=None)[source]#

Class to represent a colour map.

It can be created or selected from the built-in colour map. Please refer to the ColorMapType enum class to view all available colormaps.

Convention:: colors has shape (3, num_colors) in RGB channel order. Built-in names are case-insensitive; a custom base is an ordered list of RGB triples and is linearly resampled to num_colors.

Parameters:

base (Union[list[List[float]], str, ColorMapType]) – A list of RGB colors to define a new custom colormap or the name of a built-in colormap as str or using ColorMapType class.
num_colors (int, optional) – Number of colors in the colormap. Default: 64
device (Optional[device], optional) – The device to put the generated colormap on. Default: None
dtype (Optional[dtype], optional) – The data type of the generated colormap. Defaults to torch.get_default_dtype(). Default: None

Returns:

An object of the colormap with the num_colors length.

Examples

>>> ColorMap(base='viridis', num_colors=8).colors
tensor([[0.2813, 0.2621, 0.2013, 0.1505, 0.1210, 0.2463, 0.5259, 0.8557],
        [0.0842, 0.2422, 0.3836, 0.5044, 0.6258, 0.7389, 0.8334, 0.8886],
        [0.4072, 0.5207, 0.5543, 0.5574, 0.5334, 0.4519, 0.2880, 0.0989]])

Create a color map from the first color (RGB with range[0-1]) to the last one with num_colors length.

>>> ColorMap(base=[[0., 0.5 , 1.0], [1., 0.5, 0.]], num_colors=8).colors
tensor([[0.0000, 0.0000, 0.1250, 0.3750, 0.6250, 0.8750, 1.0000, 1.0000],
        [0.5000, 0.5000, 0.5000, 0.5000, 0.5000, 0.5000, 0.5000, 0.5000],
        [1.0000, 1.0000, 0.8750, 0.6250, 0.3750, 0.1250, 0.0000, 0.0000]])

kornia.color.RGBColor#: alias of List[float]

Color maps available:

class kornia.color.ColorMapType(value)[source]#

An enumeration for available colormaps.

List of available colormaps:

autumn = 1#

bone = 2#

jet = 3#

winter = 4#

rainbow = 5#

ocean = 6#

summer = 7#

spring = 8#

cool = 9#

hsv = 10#

brg = 11#

pink = 12#

hot = 13#

plasma = 14#

viridis = 15#

cividis = 16#

twilight = 17#

turbo = 18#

seismic = 19#

classmethod list()[source]#

Return a list of names of enumeration members.

Return type:: list[str]
Returns:: A list containing the names of enumeration members.