Modelling Consumer's Cleanliness and Whiteness Perceptions

Background To model and estimate consumer visual assessment of washing results in real laundry scenarios is nowadays an important challenge for soap and detergents R&D activities. This is obvious due to the enormous pressure that the Industry has to develop consumer values at high pace and low costs. Consumer perception of cleanliness and whiteness at the end of the washing process have been, the centre of attention of several research groups during the last 80 years inside finished products manufacturers, raw material suppliers, and independent research institutions around the world. In the last 30 years there has been several remarkable achievements in the fields of photonics, vision science, computer hardware and software that changed what is now possible to do so as to understand, model and simulate consumer visual perception under real market scenarios. Specifically, recent findings based on non-contact spectroscopy radiance measurements of real scenes, as well as international acceptance of appearance models and new affordable computer and software capabilities, have opened the door of the virtual world to conduct research and development with astonished potentials on reducing the pace and the cost of consumer-relevant innovations.

A. Modelling Cleanliness Perception Cleanliness perception is linked on the one hand to consumer ability to detect visually any remanent stained area on a fabric, in relation with some specific surround, all under some specific visual scenario, and on the other, to consumer preference of what perception is possible to be accepted as clean. That consumer ability and preference depends on a visual detection of colour differences between two areas on some washed fabric, for that reason historically, the understanding of cleanliness perception has been linked to the field of colour difference. Starting with the concept of "line element" of Helmholzt in the nineteen century, the mathematical representation of the colour differences between two surfaces has a long history, which include MacAdam ellipsoids in the 40's, numerous formulas during the 50's to 60's, and the 70's landmark: the DELTAE formula based on the CIELAB colour space. Since then several optimisation have been proposed in the colouration industry including the nearly recent DECMC and DE2000. Last advancements are based on more uniform colour spaces that correct important deficiencies of CIELAB, and have proved as of high relevancy to detergency studies. EXAMPLES A.1. DELTA-E MODEL BASED ON CIELAB GREEN SURFACE, equi-colour-difference surface, DECIELAB76 A.2. DELTA-E MODELS BASED ON CIECAM02 GREEN SURFACE, equi-colour-difference surface (sphere), based on the uniform colour space J' a' b' GREEN SURFACE, equi-colour-difference surface (sphere), based on the uniform colour space J' a' b' VIOLET SURFACE, equi-colour-difference surface (tilted ellipsoid) of a nonlinear model, consumer-relevant, based on the uniform colour space J' a' b'

B. Modelling Whiteness Perception Whiteness, a complex relative perceptual attribute, is linked on the one hand to consumer ability to perceive surface's more basics attributes like lightness, red-green opponency and yellow-blue opponency, all under some specific visual scenario, and on the other, to consumer preference of what is considered as more white, which has roots on habits and culture. The evolution of whiteness models are linked to the evolution of the level of understanding of basic and relative perceptual attributes of colours like brightness, lightness, colourfulness, chroma, saturation and hue as well as phenomenon like chromatic adaptation and colour constancy, on the consumer side, and of light reflection and fluorescence, on the surface side. Starting with the nonlinear whiteness formula of MacAdam-Judd in the 30's based on brightness and purity of dingy fabrics, whiteness models were rare in the first half of the century. The general use of fluorescent agents as whitening agents opened up a renovated interest in the topic during the 50's, 60's, finalising with the international acceptance in the 70's of the simple linear model developed by Ganz based on one tristimuli value and the chromaticity coordinates. The general acceptance of the CIELAB colour spaces during the 80's and 90's and the new uniform colour spaces based on appearance models have constituted the foundation for recent attempts to more consumer-correlated nonlinear models. EXAMPLES B.1. GANZ LINEAR MODEL GREEN PLANE, equi-whiteness plane or isoleukai of linear CIE whiteness formula, based on coordinates Y, x, y B.2. MODELS BASED ON CIECAM02 RED CURVED PLANE, equi-whiteness surface (nearly a quadric sheet) or isoleukai of the GANZ linear whiteness model, but on the uniform colour space J' a' b' VIOLET CURVED PLANE, equi-whiteness surface (quadric sheet) or isoleukai of a nonlinear model equivalent to conventional GANZ linear whiteness model, but based on the uniform colour space J' a' b' GREEN SURFACE, equi-whiteness surface (tilted two sheet elliptic hyperboloid) or isoleukai of a nonlinear model, consumer-relevant, based on the uniform colour space J' a' b'

Features A combination of an analytical and geometric modelling procedures to shape quadric functions like: tilted ellipsoids and tilted two-sheets elliptic hyperboloids capable to reach levels of correlation or more than 95% with consumer responses collected at real visual scenarios. Robust flexible procedures to obtain mathematical models that represent different consumer responses in a world wide basis, quickly and very cost effectively. It is a ready computed-based technology that can quickly be incorporated into a bigger development project within a large organisation.