A Brief History of Color yMeasurement in Tomato

David Slaughter University of California, Davisy

Window withwith Narrow Opening

Red

Glass PrismOrangeYellowGreenBlueBlueViolet

Electromagnetic Radiation

• We use a ruler to measure length, g ,and a scale to measure mass. I ’t th t• Isn’t there a way to measure color?

A Tale of Two CitiesRi h d HRichard Hunter & USDA

PTAB & MagnusonEngineers, Inc.

Karl Ewald Hering(1834 1918)(1834-1918)

German Physiologist• When viewing a mixture of psychologically

pure red and pure green lights,a person will see: red• a person will see: red, green, or white,

• never yellow.y

• Hering created a theory of l i i b d th

Humancolor vision based on three opposing pairs:

• Red – Green

Eye

• Yellow – Blue, and • Black - White

Afterimage Example• Stare for ~30s at center, then

switch to white slide.

Richard S. HunterOptical EngineerOptical Engineer

• In 1948, he developed the Hunter L, a, b color space

– Based upon the Hering’s p gred-green, yellow-blue opponent color axes

• Hunter’s Objective:– Create a “Uniform”

color space• A one-unit change in a or

b is consistent with humanb is consistent with human perception of the amount of color change.

Hunter L, a, b color spaceSi l i th• Since color is the principal attribute that

iconsumers use in purchasing decisions, th H t & b lthe Hunter a & b color scores were adopted b th i d tby the industry as a measure of quality.

– USDA Processed Tomato Juice Color ScoreUSDA Color = 25.715 +K1 * [a –K2*b] K1 = 0.956, K2 = 1.828

S. R. Whipple, 1955. COLOR INSPECTION—FRESH FRUITS AND VEGETABLES

• First electronic instrument to measure tomato ripeness was developed in the early 1950’sdeveloped in the early 1950 s.

• Designed to measure the color of the surface of two halves of a t ttomato.

• Color score was the G/R ratio of green reflectance (546nm) to red reflectance (640 nm).

• Initially used to train inspectors and to determine the grade ofand to determine the grade of “doubtful” tomatoes.

• Modified in the 1970’s to measure the color of demeasure the color of de-aerated juice.

Measuring T t RiTomato Ripeness

reen

ker

Red R

ed

• Kader & Morris, 1978

atur

e G

Bre

ak

Pin

kLi

ght R

– Used an Agtron E5-W Ma

urni

ng

• Agtron score was well correlated with ripeness

Tu

correlated with ripeness stage.

Agtron E5-MC l i P i T t I tiColor in Processing Tomato Inspection

Red Neon &Red Neon & Green Mercury LampsLamps

Tomato Juice Reflectance35

30 20

PTAB Color

(502

nm

)

25

(%)

26 31 33 so

rban

ce

15

20

ctan

ce 33

35 43

open

e A

bs

banc

e

10

Ref

lec

Lyco

yll A

bsor

b

70 n

m)

0

5

Chl

orop

hy (67

0 400 450 500 550 600 650 700

Wavelength (nm)

C

INTERINTER--INSTRUMENT AGREEMENT OF AGTRON & LED INSTRUMENT AGREEMENT OF AGTRON & LED MACHINES READING A SINGLE TOMATO JUICE SAMPLEMACHINES READING A SINGLE TOMATO JUICE SAMPLE

100LEDAGTRONR

ER

E

80AGTRON

HIN

ES

HIN

ES

LO

R S

CO

LO

R S

CO

60

OF

MA

CH

OF

MA

CH

ME

CO

LM

E C

OL

20

40

RC

EN

T O

RC

EN

T O

TH

E S

AM

TH

E S

AM

0

20

PER

PER

GIV

ING

G

IVIN

G

024 25 26 27 28 29 30 31 32

COLOR SCORECOLOR SCORE

UC Davis LED Tomato Color MeterTEMPERATURECONTROLLEDCHAMBER (110 oF)

LED LAMP ARRAYCHAMBER (110 oF)

REFLECTEDLIGHTLIGHT

DETECTOR

TOMATO JUICE

CIE L*, a*, b* color system• “Uniform” color

systemsystem– Developed in

1976 by the yInternational Commission on Illumination

– “Improved”i f H tversion of Hunter

L, a, b color space

Hunter L,a,b vs. CIE L*, a*, b*Hunter L,a,b vs. CIE L , a , b

HUNTER CIE1

L 10 Y L* 116 * YY

1316

a 17.5 * 1.02X Y

Yn

X 13 Y

13

a

Y

7* Y 0 85Z a* 500 * X

Xn

3

YYn

3

b 7* Y 0.85Z

Y

b* 200 * Y

13 Z

13

b 200 *

Yn

Zn

Comparison of Color Space Uniformity

Hunter Lab CIE L*a* b*

Albert MunsellAmerican PainterAmerican Painter

1915The original Munsell

system is almost f tl if

• Hue = name of color(e g purple green yellow)

perfectly uniform.

– (e.g., purple, green, yellow)

• Value, (or lightness)Divided into 11 equal steps– Divided into 11 equal steps

• (black=0, white = 10)

– Value ~ √ave reflectanceValue √ave. reflectance

• Chroma, purity, or saturation

Konica Minolta CR-10

CIE L* C* H*CIE L ,C H Color Meter

• Over 13,000 peaches evaluated for maturity with the CR-10 in 2006.

• Operated by SPI p yinspectors in 2006.

Hue Color Wheel80

45

90

8080 degrees

45

0180

270Standard Illuminant: D65Standard Illuminant: D6510 Degree Observer

UC Davis

PTAB Color Situation• The current LED technology is

beginning to require increasedbeginning to require increased maintenance and will require replacement.

• The opportunity exists to make a significant change in how maturity/color is determined at har estdetermined at harvest.

PTAB Color Proposal for 2012• PTAB would start

using the Minolta CR-410 Colorimeter for

l i i icolor inspection in 2012.

• CIE H* (Hue Angle) would become the

ffofficial grade.• Hunter L, a, b, and• CIE L*, a*, b* could

be provided at no

PTAB Color vs. USDA Color

45

UC Davis 2006 Color Study Results

35

40

co

re

30

35

Co

lor

Sc

25

PTA

B C

15

20

10 20 30 40 50Hunter / USDA Processed K1*(a - K2*b) Color Score

PTAB Color vs. Hue Angle

45

ore

UC Davis 2006 Color Study Results

35

40

Co

lor

Sc

30

35

d J

uic

e C

25

Sc

ree

ne

d

15

20

PTA

B S

10 20 30 40 50CIE H* (Hue Angle) Unscreened Juice

PTAB Color vs. L*/a* Ratio

45

ore

UC Davis 2006 Color Study Results

35

40

Co

lor

Sc L* ≈ Green

a* ≈ Red

30

35

d J

uic

e C

L*/a* ≈ G/R = Agtron

25

Sc

ree

ne

d

15

20

PTA

B S

0 0.5 1 1.5 2 2.5 3CIE L*/a* Unscreened Juice

2011 Prototype Flow-Through Color Grading SystemGrading System

• Blends juice sample,• Measures Color, pH,

and Soluble Solids,• Self-cleaning.