colour.recovery.RGB_to_msds_Smits1999

colour.recovery.RGB_to_msds_Smits1999(RGB: ArrayLike, basis: MultiSpectralDistributions | None = None) → NDArrayFloat

Recover spectral values from RGB colourspace array using the Smits (1999) decomposition algorithm.

This is a vectorised implementation supporting multi-dimensional arrays.

Parameters:

• RGB (ArrayLike) – RGB colourspace array to recover spectral values from. The last dimension must be size 3.

• basis (MultiSpectralDistributions | None) – Multi-spectral distributions basis with signals: white, cyan, magenta, yellow, red, green, blue. Defaults to MSDS_SMITS1999.

Returns:

Recovered spectral values with shape (*RGB.shape[:-1], wavelengths).

Return type:

numpy.ndarray

Notes

Domain	Scale - Reference	Scale - 1
RGB	1	1

References

[Smi99]

Examples

>>> import numpy as np
>>> RGB = np.array(
...     [
...         [0.45623196, 0.03080455, 0.04093343],
...         [0.05438271, 0.29877169, 0.07188444],
...         [0.01863137, 0.05139773, 0.28887675],
...     ]
... )
>>> RGB_to_msds_Smits1999(RGB).shape
(3, 10)
>>> float(RGB_to_msds_Smits1999(RGB)[0, 0])
0.0829...