New luster formula for the characterization of hair tresses using polarization imaging

Abstract

Hair luster is one of the most important parameters of visual appearance perceived by consumers. Current luster formulae (TRI, Reich-Robbins, ...) are optimized for goniophotometric measurements. They are based on a mathematical decomposition of reflected light into specular and diffused light and the meaurement of the shine peak width on the fitted angular distributions. In this expose, we are describing a polarization imaging system measuring luster of hair tresses with an innovative algorithm. Using polarization imaging allows to physically separating the specular light from the diffused light for each pixel of the imaged tress. Angular distributions of the specular and diffused light are obtained in a few seconds. Where conventional methods calculate the shine peak width on the angular distribution, the imaging system imitates the human eye and calculates the shine width directly on the image. The new formula combines different measured parameters to objectively quantify luster. It was designed to exhibit a higher correlation with visual perception along with a higher sensitivity. Results obtained with conventional formulae are compared on different hair tresses, treated and untreated. The new formula is found to be consistent for a whole range of hair colors, from light to dark.