Evaluating a Multi-Spectral Imaging System for Mapping Pigments in Human Skin

A multi-spectral imaging system is evaluated for mapping melanin density, total-hemoglobin density, and oxygen saturation in human skin. In this system, the distribution of pigments in human skin is estimated and displayed from digital video signals using three pre-computed “look-up” tables for color conversions. The accuracy of the system is analyzed based on computer simulation by changing the number of bands, quantization levels of the camera system, and dimensions of approximation for spectral reflectance. The optimal system is examined based on the results of computer simulation.