Characterization of the human iris spectral reflectance with a multispectral imaging system

Area of Science:

• Ophthalmology
• Biomedical Optics
• Color Science

Background:

• Accurate measurement of human iris spectral reflectance and color is crucial for clinical applications like prosthetics and cosmetic contact lenses.
• Existing methods may lack the precision or efficiency required for comprehensive iris characterization.

Purpose of the Study:

• To develop and optimize a multispectral system for precise measurement of human iris spectral reflectance and color.
• To evaluate the efficacy of different filter sets, reconstruction algorithms, and training datasets for accurate color reproduction.

Main Methods:

• Designed and tested a multispectral imaging system with various filter sets as acquisition channels.
• Analyzed multiple spectral reconstruction algorithms and utilized diverse sample sets for training.
• Validated the system's accuracy using color difference metrics (CIEDE2000) and root mean square errors.

Main Results:

• A conventional three-channel RGB camera system was found to be sufficient for reconstructing iris spectral reflectances with high accuracy.
• Achieved average color differences of approximately 2-3 CIEDE2000 units and root mean square errors around 0.01.
• Successfully characterized 100 real irises, 68 clinical prostheses, and 17 cosmetic contact lenses.

Conclusions:

• The developed multispectral system, utilizing a standard RGB camera, provides accurate spectral reflectance and color measurements of the human iris.
• The system is suitable for characterizing biological irises, prosthetic devices, and cosmetic contact lenses.
• This technology offers a reliable tool for research, clinical practice, and the cosmetic industry.