Meritxell Vilaseca,1,*
Rita Mercadal,1
Jaume Pujol,1
Monserrat Arjona,1
Marta de Lasarte,1
Rafael Huertas,2
Manuel Melgosa,2
and Francisco H. Imai3
1Centre for Sensors, Instruments and Systems Development (CD6), Technical University of Catalonia (UPC), Rambla Sant Nebridi 10, Terrassa, Barcelona 08222, Spain
2Department of Optics, University of Granada (UGR), Campus Fuentenueva s/n, Granada 18071, Spain
3Samsung Information Systems America (SISA), 75 West Plumeria Drive, San Jose, California 95134, USA
Meritxell Vilaseca, Rita Mercadal, Jaume Pujol, Monserrat Arjona, Marta de Lasarte, Rafael Huertas, Manuel Melgosa, and Francisco H. Imai, "Characterization of the human iris spectral reflectance with a multispectral imaging system," Appl. Opt. 47, 5622-5630 (2008)
We present a multispectral system developed and optimized for measurement of the spectral reflectance and the color of the human iris. We tested several sets of filters as acquisition channels, analyzed different reconstruction algorithms, and used different samples as training sets. The results obtained show that a conventional three-channel color camera (RGB) was enough to reconstruct the analyzed reflectances with high accuracy, obtaining averaged color differences of around 2–3 CIEDE2000 units and root mean square errors of around 0.01. The device developed was used to characterize 100 real irises corresponding to 50 subjects, 68 prostheses used in clinical practice, and 17 cosmetic colored contact lenses.
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
Color Difference and Root Mean Square Error (Mean ± St. Dev.) for the 52 Human Irisesa
PSE
PCA
Configuration 1
CIEDE2000
RMSE
CIEDE2000
RMSE ()
3 channels
6 channels
Configuration 2
3 channels
7 channels
Training, 52 human irises; test, 52 human irises.
Table 3
Color Difference and Root Mean Square Error (Mean ± St. Dev.) for the 52 Human Irisesa
Training
CIEDE2000
RMSE
28 irises
16 irises
Test, 52 human irises; method, PSE; configuration, 1, with three channels.
Table 4
Color Difference and Root Mean Square Error (Mean ± St. Dev.) for the 34 Brown Irises and the 18 Blue–Green Irisesa
CIEDE2000
RMSE
Brown eyes
Blue-green eyes
Training, 34 brown irises/18 brown irises, respectively; test, 34 brown irises/18 brown irises, respectively; method, PSE; configuration, 1, with three channels.
Table 5
Cumulative Proportion Index (%) for the First Three Principal Components (PC) Corresponding to the 100 Human Irises, 68 Prostheses, and 17 Colored Contact Lenses
PC
Irises
Prostheses
Contact Lenses
1
64.71
68.85
51.71
2
95.82
97.35
94.67
3
99.21
99.08
99.01
Table 6
Color Difference and Root Mean Square Error (Mean, Standard Deviation, Minimum, and Maximum Values) for the 100 Human Irises, 68 Prostheses, and 17 Colored Contact Lensesa
CIEDE2000
RMSE
, human irises
Mean
2.46
0.0108
St. Dev.
1.10
6.4739e-3
Min.
0.50
2.3433e-3
Max.
5.44
0.0345
, prostheses
Mean
1.97
0.0111
St. Dev.
0.91
6.2001e-3
Min.
0.46
2.4249e-3
Max.
4.44
0.0342
, contact lenses
Mean
2.57
0.0105
St. Dev.
0.95
6.1035e-3
Min.
0.77
2.5625e-3
Max.
4.01
0.0229
; method, PSE; configuration, 1, with three channels.
Table 7
Color Difference and Root Mean Square Error (Mean, Standard Deviation, Minimum, and Maximum Values) for the 100 Human Irises, 68 Prostheses, and 17 Colored Contact Lenses
CIEDE2000
RMSE
Training set, human irises; test set, prostheses
Mean
2.90
0.0179
St. Dev.
1.22
0.0115
Min.
0.57
3.0910e-3
Max.
5.97
0.0569
Training set, human irises; test set, contact lenses
Mean
2.42
0.0129
St. Dev.
1.09
6.8951e-3
Min.
0.79
2.3770e-3
Max.
5.48
0.0316
Training set, prostheses; test set, human irises
Mean
3.78
0.0165
St. Dev.
1.23
8.4533e-3
Min.
0.55
4.6276e-3
Max.
7.58
0.0569
Training set, prostheses; test set, contact lenses
Mean
4.04
0.0182
St. Dev.
1.05
7.0793e-3
Min.
1.82
9.6825e-3
Max.
6.78
0.0390
Training set, contact lenses; test set, human irises
Mean
3.79
0.0167
St. Dev.
1.77
9.0554e-3
Min.
0.47
1.9612e-3
Max.
8.43
0.0458
Training set, contact lenses; test set, prostheses
Mean
5.86
0.0250
St. Dev.
2.34
0.0118
Min.
1.30
5.3829e-3
Max.
11.38
0.0573
; method, PSE; configuration, 1, with three channels.
Tables (7)
Table 1
Color Difference and Root Mean Square Error (Mean ± St. Dev.) for the 52 Human Irisesa
Color Difference and Root Mean Square Error (Mean ± St. Dev.) for the 52 Human Irisesa
PSE
PCA
Configuration 1
CIEDE2000
RMSE
CIEDE2000
RMSE ()
3 channels
6 channels
Configuration 2
3 channels
7 channels
Training, 52 human irises; test, 52 human irises.
Table 3
Color Difference and Root Mean Square Error (Mean ± St. Dev.) for the 52 Human Irisesa
Training
CIEDE2000
RMSE
28 irises
16 irises
Test, 52 human irises; method, PSE; configuration, 1, with three channels.
Table 4
Color Difference and Root Mean Square Error (Mean ± St. Dev.) for the 34 Brown Irises and the 18 Blue–Green Irisesa
CIEDE2000
RMSE
Brown eyes
Blue-green eyes
Training, 34 brown irises/18 brown irises, respectively; test, 34 brown irises/18 brown irises, respectively; method, PSE; configuration, 1, with three channels.
Table 5
Cumulative Proportion Index (%) for the First Three Principal Components (PC) Corresponding to the 100 Human Irises, 68 Prostheses, and 17 Colored Contact Lenses
PC
Irises
Prostheses
Contact Lenses
1
64.71
68.85
51.71
2
95.82
97.35
94.67
3
99.21
99.08
99.01
Table 6
Color Difference and Root Mean Square Error (Mean, Standard Deviation, Minimum, and Maximum Values) for the 100 Human Irises, 68 Prostheses, and 17 Colored Contact Lensesa
CIEDE2000
RMSE
, human irises
Mean
2.46
0.0108
St. Dev.
1.10
6.4739e-3
Min.
0.50
2.3433e-3
Max.
5.44
0.0345
, prostheses
Mean
1.97
0.0111
St. Dev.
0.91
6.2001e-3
Min.
0.46
2.4249e-3
Max.
4.44
0.0342
, contact lenses
Mean
2.57
0.0105
St. Dev.
0.95
6.1035e-3
Min.
0.77
2.5625e-3
Max.
4.01
0.0229
; method, PSE; configuration, 1, with three channels.
Table 7
Color Difference and Root Mean Square Error (Mean, Standard Deviation, Minimum, and Maximum Values) for the 100 Human Irises, 68 Prostheses, and 17 Colored Contact Lenses
CIEDE2000
RMSE
Training set, human irises; test set, prostheses
Mean
2.90
0.0179
St. Dev.
1.22
0.0115
Min.
0.57
3.0910e-3
Max.
5.97
0.0569
Training set, human irises; test set, contact lenses
Mean
2.42
0.0129
St. Dev.
1.09
6.8951e-3
Min.
0.79
2.3770e-3
Max.
5.48
0.0316
Training set, prostheses; test set, human irises
Mean
3.78
0.0165
St. Dev.
1.23
8.4533e-3
Min.
0.55
4.6276e-3
Max.
7.58
0.0569
Training set, prostheses; test set, contact lenses
Mean
4.04
0.0182
St. Dev.
1.05
7.0793e-3
Min.
1.82
9.6825e-3
Max.
6.78
0.0390
Training set, contact lenses; test set, human irises
Mean
3.79
0.0167
St. Dev.
1.77
9.0554e-3
Min.
0.47
1.9612e-3
Max.
8.43
0.0458
Training set, contact lenses; test set, prostheses
Mean
5.86
0.0250
St. Dev.
2.34
0.0118
Min.
1.30
5.3829e-3
Max.
11.38
0.0573
; method, PSE; configuration, 1, with three channels.