vsl-ial 0.0.7

Color spaces made reasonable

VSL Image Appearance Library

Python module for working with color spaces and stuff

Usage

Color space conversion:

• Convert from sRGB to libRGB

from vsl_ial.cs import convert, sRGB, linRGB

convert(sRGB(), linRGB(), color=[0.12412, 0.07493, 0.3093])

Available color spaces

Function name	Color space description
XYZ()	CIE XYZ
CIExyY()	CIE xyY
CIELUV()	CIELUV
CIELAB()	CIELAB
ProLab()	proLab
Oklab()	Oklab
ICaCb()	ICaCb
ICtCp()	ICtCp
JzAzBz()	Jzazbz
CAM02LCD()	CAM02-LCD
CAM02SCD()	CAM02-SCD
CAM02UCS()	CAM02-UCS
CAM16LCD()	CAM16-LCD
CAM16SCD()	CAM16-SCD
CAM16UCS()	CAM16-UCS
sRGB()	sRGB
linRGB()	linear sRGB
LMS()	LMS
Opponent()	Opponent color space by Zhang and Wandell
PCS23-UCS()	Uniform Color Space with Advanced Hue Linearity

Available measures of the consistency between perceived and computed color differences

• CV
• PF/3
• STRESS
• Mean STRESS
• Group STRESS
• Weighted group STRESS

Model evaluation

Run

python -m vsl_ial.eval

To get uniformity table

	COMBVD	BFD-P d65	BFD-P m	BFD-P c	RIT-DuPont	Witt	Leeds	Munsell-3.1.0*
CAM16-SCD	0.295	0.254	0.346	0.283	0.237	0.306	0.219	0.0871
CAM16-UCS	0.305	0.271	0.35	0.297	0.206	0.31	0.245	0.0883
CAM16-LCD	0.339	0.311	0.372	0.366	0.214	0.372	0.292	0.107
CAM02-SCD	0.296	0.266	0.338	0.303	0.244	0.303	0.221	0.111
CAM02-UCS	0.306	0.28	0.343	0.321	0.213	0.305	0.246	0.116
CAM02-LCD	0.339	0.318	0.366	0.404	0.223	0.366	0.296	0.169
PCS23-UCS	0.311	0.289	0.325	0.379	0.3	0.381	0.332	0.0741
CIELAB	0.426	0.41	0.433	0.543	0.334	0.517	0.401	0.281
ProLab	0.441	0.451	0.429	0.485	0.302	0.519	0.394	0.177
Oklab	0.471	0.515	0.424	0.416	0.318	0.452	0.45	0.074
JzAzBz	0.418	0.404	0.424	0.494	0.385	0.474	0.451	0.127
ICaCb	0.391	0.396	0.38	0.424	0.248	0.474	0.373	0.14
ICtCp	0.463	0.481	0.441	0.636	0.423	0.559	0.397	0.228
* subset of Munsell dataset

For module developers

• How to make whl file:

python -m pip install --upgrade build
python -m build

• How to run unit tests:

python -m unittest

• How to run coverage:

python -m coverage run -m unittest
python -m coverage html