Source code for colour.appearance.rlab

"""
RLAB Colour Appearance Model
============================

Defines the *RLAB* colour appearance model objects:

-   :attr:`colour.VIEWING_CONDITIONS_RLAB`
-   :attr:`colour.D_FACTOR_RLAB`
-   :class:`colour.CAM_Specification_RLAB`
-   :func:`colour.XYZ_to_RLAB`

References
----------
-   :cite:`Fairchild1996a` : Fairchild, M. D. (1996). Refinement of the RLAB
    color space. Color Research & Application, 21(5), 338-346.
    doi:10.1002/(SICI)1520-6378(199610)21:5<338::AID-COL3>3.0.CO;2-Z
-   :cite:`Fairchild2013w` : Fairchild, M. D. (2013). The RLAB Model. In Color
    Appearance Models (3rd ed., pp. 5563-5824). Wiley. ISBN:B00DAYO8E2
"""

from __future__ import annotations

from dataclasses import dataclass, field

import numpy as np

from colour.algebra import matrix_dot, sdiv, sdiv_mode, spow, vector_dot
from colour.appearance.hunt import MATRIX_XYZ_TO_HPE, XYZ_to_rgb
from colour.hints import ArrayLike, NDArrayFloat
from colour.utilities import (
    CanonicalMapping,
    MixinDataclassArray,
    as_float,
    as_float_array,
    from_range_degrees,
    row_as_diagonal,
    to_domain_100,
    tsplit,
)

__author__ = "Colour Developers"
__copyright__ = "Copyright 2013 Colour Developers"
__license__ = "BSD-3-Clause - https://opensource.org/licenses/BSD-3-Clause"
__maintainer__ = "Colour Developers"
__email__ = "colour-developers@colour-science.org"
__status__ = "Production"

__all__ = [
    "MATRIX_R",
    "VIEWING_CONDITIONS_RLAB",
    "D_FACTOR_RLAB",
    "CAM_ReferenceSpecification_RLAB",
    "CAM_Specification_RLAB",
    "XYZ_to_RLAB",
]

MATRIX_R: NDArrayFloat = np.array(
    [
        [1.9569, -1.1882, 0.2313],
        [0.3612, 0.6388, 0.0000],
        [0.0000, 0.0000, 1.0000],
    ]
)
"""*RLAB* colour appearance model precomputed helper matrix."""

VIEWING_CONDITIONS_RLAB: CanonicalMapping = CanonicalMapping(
    {"Average": 1 / 2.3, "Dim": 1 / 2.9, "Dark": 1 / 3.5}
)
VIEWING_CONDITIONS_RLAB.__doc__ = """
Reference *RLAB* colour appearance model viewing conditions.

References
----------
:cite:`Fairchild1996a`, :cite:`Fairchild2013w`
"""

D_FACTOR_RLAB: CanonicalMapping = CanonicalMapping(
    {
        "Hard Copy Images": 1,
        "Soft Copy Images": 0,
        "Projected Transparencies, Dark Room": 0.5,
    }
)
D_FACTOR_RLAB.__doc__ = """
*RLAB* colour appearance model *Discounting-the-Illuminant* factor values.

References
----------
:cite:`Fairchild1996a`, :cite:`Fairchild2013w`

Aliases:

-   'hard_cp_img': 'Hard Copy Images'
-   'soft_cp_img': 'Soft Copy Images'
-   'projected_dark': 'Projected Transparencies, Dark Room'
"""
D_FACTOR_RLAB["hard_cp_img"] = D_FACTOR_RLAB["Hard Copy Images"]
D_FACTOR_RLAB["soft_cp_img"] = D_FACTOR_RLAB["Soft Copy Images"]
D_FACTOR_RLAB["projected_dark"] = D_FACTOR_RLAB["Projected Transparencies, Dark Room"]


@dataclass
class CAM_ReferenceSpecification_RLAB(MixinDataclassArray):
    """
    Define the *RLAB* colour appearance model reference specification.

    This specification has field names consistent with *Fairchild (2013)*
    reference.

    Parameters
    ----------
    LR
        Correlate of *Lightness* :math:`L^R`.
    CR
        Correlate of *achromatic chroma* :math:`C^R`.
    hR
        *Hue* angle :math:`h^R` in degrees.
    sR
        Correlate of *saturation* :math:`s^R`.
    HR
        *Hue* :math:`h` composition :math:`H^R`.
    aR
        Red-green chromatic response :math:`a^R`.
    bR
        Yellow-blue chromatic response :math:`b^R`.

    References
    ----------
    :cite:`Fairchild1996a`, :cite:`Fairchild2013w`
    """

    LR: float | NDArrayFloat | None = field(default_factory=lambda: None)
    CR: float | NDArrayFloat | None = field(default_factory=lambda: None)
    hR: float | NDArrayFloat | None = field(default_factory=lambda: None)
    sR: float | NDArrayFloat | None = field(default_factory=lambda: None)
    HR: float | NDArrayFloat | None = field(default_factory=lambda: None)
    aR: float | NDArrayFloat | None = field(default_factory=lambda: None)
    bR: float | NDArrayFloat | None = field(default_factory=lambda: None)


[docs] @dataclass class CAM_Specification_RLAB(MixinDataclassArray): """ Define the *RLAB* colour appearance model specification. This specification has field names consistent with the remaining colour appearance models in :mod:`colour.appearance` but diverge from *Fairchild (2013)* reference. Parameters ---------- J Correlate of *Lightness* :math:`L^R`. C Correlate of *achromatic chroma* :math:`C^R`. h *Hue* angle :math:`h^R` in degrees. s Correlate of *saturation* :math:`s^R`. HC *Hue* :math:`h` composition :math:`H^C`. a Red-green chromatic response :math:`a^R`. b Yellow-blue chromatic response :math:`b^R`. Notes ----- - This specification is the one used in the current model implementation. References ---------- :cite:`Fairchild1996a`, :cite:`Fairchild2013w` """ J: NDArrayFloat | None = field(default_factory=lambda: None) C: NDArrayFloat | None = field(default_factory=lambda: None) h: NDArrayFloat | None = field(default_factory=lambda: None) s: NDArrayFloat | None = field(default_factory=lambda: None) HC: NDArrayFloat | None = field(default_factory=lambda: None) a: NDArrayFloat | None = field(default_factory=lambda: None) b: NDArrayFloat | None = field(default_factory=lambda: None)
[docs] def XYZ_to_RLAB( XYZ: ArrayLike, XYZ_n: ArrayLike, Y_n: ArrayLike, sigma: ArrayLike = VIEWING_CONDITIONS_RLAB["Average"], D: ArrayLike = D_FACTOR_RLAB["Hard Copy Images"], ) -> CAM_Specification_RLAB: """ Compute the *RLAB* model color appearance correlates. Parameters ---------- XYZ *CIE XYZ* tristimulus values of test sample / stimulus. XYZ_n *CIE XYZ* tristimulus values of reference white. Y_n Absolute adapting luminance in :math:`cd/m^2`. sigma Relative luminance of the surround, see :attr:`colour.VIEWING_CONDITIONS_RLAB` for reference. D *Discounting-the-Illuminant* factor normalised to domain [0, 1]. Returns ------- CAM_Specification_RLAB *RLAB* colour appearance model specification. Notes ----- +------------+-----------------------+---------------+ | **Domain** | **Scale - Reference** | **Scale - 1** | +============+=======================+===============+ | ``XYZ`` | [0, 100] | [0, 1] | +------------+-----------------------+---------------+ | ``XYZ_n`` | [0, 100] | [0, 1] | +------------+-----------------------+---------------+ +------------------------------+-----------------------\ +---------------+ | **Range** | **Scale - Reference** \ | **Scale - 1** | +==============================+=======================\ +===============+ | ``CAM_Specification_RLAB.h`` | [0, 360] \ | [0, 1] | +------------------------------+-----------------------\ +---------------+ References ---------- :cite:`Fairchild1996a`, :cite:`Fairchild2013w` Examples -------- >>> XYZ = np.array([19.01, 20.00, 21.78]) >>> XYZ_n = np.array([109.85, 100, 35.58]) >>> Y_n = 31.83 >>> sigma = VIEWING_CONDITIONS_RLAB["Average"] >>> D = D_FACTOR_RLAB["Hard Copy Images"] >>> XYZ_to_RLAB(XYZ, XYZ_n, Y_n, sigma, D) # doctest: +ELLIPSIS CAM_Specification_RLAB(J=49.8347069..., C=54.8700585..., \ h=286.4860208..., s=1.1010410..., HC=None, a=15.5711021..., \ b=-52.6142956...) """ XYZ = to_domain_100(XYZ) XYZ_n = to_domain_100(XYZ_n) Y_n = as_float_array(Y_n) D = as_float_array(D) sigma = as_float_array(sigma) # Converting to cone responses. LMS_n = XYZ_to_rgb(XYZ_n) # Computing the :math:`A` matrix. LMS_l_E = 3 * LMS_n / np.sum(LMS_n, axis=-1)[..., None] LMS_p_L = (1 + spow(Y_n[..., None], 1 / 3) + LMS_l_E) / ( 1 + spow(Y_n[..., None], 1 / 3) + 1 / LMS_l_E ) LMS_a_L = (LMS_p_L + D[..., None] * (1 - LMS_p_L)) / LMS_n M = matrix_dot(matrix_dot(MATRIX_R, row_as_diagonal(LMS_a_L)), MATRIX_XYZ_TO_HPE) XYZ_ref = vector_dot(M, XYZ) X_ref, Y_ref, Z_ref = tsplit(XYZ_ref) # Computing the correlate of *Lightness* :math:`L^R`. LR = 100 * spow(Y_ref, sigma) # Computing opponent colour dimensions :math:`a^R` and :math:`b^R`. aR = 430 * (spow(X_ref, sigma) - spow(Y_ref, sigma)) bR = 170 * (spow(Y_ref, sigma) - spow(Z_ref, sigma)) # Computing the *hue* angle :math:`h^R`. hR = np.degrees(np.arctan2(bR, aR)) % 360 # TODO: Implement hue composition computation. # Computing the correlate of *chroma* :math:`C^R`. CR = np.hypot(aR, bR) # Computing the correlate of *saturation* :math:`s^R`. with sdiv_mode(): sR = sdiv(CR, LR) return CAM_Specification_RLAB( LR, CR, as_float(from_range_degrees(hR)), sR, None, as_float(aR), as_float(bR), )