Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

correct implementation of the CIEDE2000 formula is non-

trivial and our goal in this article is to assist in this process.


This article has three main contributions. First, we pro-
冉 冑
G⫽0.5 1⫺
C៮ *ab7
C៮ *ab7⫹257
冊 (4)

vide a comprehensive set of data for testing software im- a⬘i⫽共1⫹G兲a*i i⫽1, 2 (5)
plementations of the CIEDE2000 color-difference equation. C⬘i⫽ 冑共a⬘i 兲 ⫹共b*i 兲
2 2
i⫽1, 2 (6)
The data are specifically designed for exercising compo-
nents in the CIEDE2000 equations that may be erroneously
0
再 b i ⫽a⬘i⫽0
*
h⬘i⫽ tan⫺1共b*, a⬘兲 otherwise
i i
i⫽1, 2 (7)
computed and are not exercised by the original worked
examples included with the draft standard. We also point 2. Calculate ⌬L⬘, ⌬C⬘, ⌬H⬘:
out several of the common errors and indicate parts of the ⌬L⬘⫽L*2⫺L*1 (8)
test data that help identify these. Second, we provide im-
plementations of the CIEDE2000 color-difference formula ⌬C⬘⫽C⬘2⫺C⬘1 (9)


in Microsoft Excel and MATLAB.5 These implementations 0 C⬘1C⬘2⫽0
agree with the original worked examples included in the h⬘2⫺h⬘1 C⬘1C⬘2⫽0; 兩h⬘2⫺h⬘1兩ⱕ180⬚
⌬h⬘⫽ 共h⬘⫺h⬘兲⫺360 C⬘C⬘⫽0; 共h⬘⫺h⬘兲⬎180⬚
CIE Technical Report and with each other on the supple- 2 1 1 2 2 1

mental data included in this article. We believe these to be 共h⬘2⫺h⬘1兲⫹360 C⬘1C⬘2⫽0; 共h⬘2⫺h⬘1兲⬍⫺180⬚
correct implementations of the equations. Third, we high- (10)
light and characterize discontinuities in the CIEDE2000
color-difference formula that arise from its defining equa-
tions.
⌬H⬘⫽2 冑C⬘1C⬘2 sin 冉 冊
⌬h⬘
2
(11)

This article addresses only the computational and math- 3. Calculate CIEDE2000 Color-Difference ⌬E 00 :
ematical aspects of the CIEDE2000 equations and does not
in any way address psychophysical evaluation of the for- L៮ ⬘⫽共L*1⫹L*2兲/2 (12)
mula, attempt to improve its uniformity, or define its do- ៮ ⬘⫽共C⬘1⫹C⬘2兲/2
C (13)
main of applicability. Some comments of that nature are


included in the notes by Kuehni6 and Luo et al.7 published h⬘1⫹h⬘2
in this journal. 兩h⬘1⫺h⬘2兩ⱕ180⬚; C⬘1C⬘2⫽0
2
h⬘1⫹h⬘2⫹360⬚
兩h⬘1⫺h⬘2兩⬎180⬚; 共h⬘1⫹h⬘2兲⬍360⬚;
THE CIEDE2000 COLOR-DIFFERENCE FORMULA 2
h៮ ⬘⫽ C⬘1C⬘2⫽0
In this section, we briefly outline the equations for the h⬘1⫹h⬘2⫺360⬚
兩h⬘1⫺h⬘2兩⬎180⬚; 共h⬘1⫹h⬘2兲ⱖ360⬚;
computation of the CIEDE2000 color difference. The de- 2
scription follows the presentation of Luo, Cui, and Rigg2 but C⬘1C⬘2⫽0
is slightly modified to make it closer to an algorithmic 共h⬘1⫹h⬘2兲 C⬘1C⬘2⫽0
statement. It is included here to relate implementation notes
(14)
and comments in subsequent sections to the equations pre-
sented here. T⫽1⫺0.17 cos共h៮ ⬘⫺30⬚兲⫹0.24 cos共2h៮ ⬘兲
The CIEDE2000 color-difference formula is based on the
CIELAB color space. Given a pair of color values in ⫹0.32 cos共3h៮ ⬘⫹6⬚兲⫺0.20 cos共4h៮ ⬘⫺63⬚兲 (15)

再冋 册冎
CIELAB space L *1 ,a *1 ,b *1 and L *2 ,a *2 ,b *2 , we denote the
h៮ ⬘⫺275⬚ 2
CIEDE2000 color difference between them as follows†: ⌬␪⫽30 exp ⫺ (16)
25
⌬E 00共L *1,a *1,b *1; L *2,a *2,b *2兲 ⫽ ⌬E 12
00 ⫽ ⌬E 00.


(1)
C៮ ⬘7
Given two CIELAB color values {L *i ,a *i ,b *i } 2i⫽1
and para- RC⫽2 ៮ 7 (17)
C⬘ ⫹257
metric weighting factors k L , k C , and k H , the process of
computation of the color difference is summarized in the 0.015共L៮ ⬘⫺50兲2
SL⫽1⫹ (18)
following equations, grouped as three main steps.
冑20⫹共L៮ ⬘⫺50兲2
1. Calculate C⬘i , h⬘i : SC⫽1⫹0.045C៮ ⬘ (19)

C*i,ab⫽ 冑共a*i 兲2⫹共b*i 兲2 i⫽1, 2 (2) SH⫽1⫹0.015C៮ ⬘T (20)


C*1,ab⫹C*2,ab
C៮ *ab⫽ (3) RT⫽⫺sin共2⌬␪ 兲 RC (21)
2
⌬E12
00⫽⌬E00共L*
1,a*
1,b*
1; L*
2,a* 2兲
2,b*

⫽ 冑冉 冊 冉 冊 冉 冊 冉 冊冉 冊
⌬L⬘ 2 ⌬C⬘ 2 ⌬H⬘ 2
kLSL

kCSC

kHSH
⫹RT
⌬C⬘ ⌬H⬘
kCSC kHSH
. (22)

You might also like