The Role of Individual Electric Fields and Intrinsic Oscillatory Brain Activity for Rhythmic Non-Invasive Brain Stimulation

Abstract

This paper presents an automatic system for modeling and controlling color quality of dyed leathers of an Italian tannery. The proposed software is implemented within the IT company system, and is fully integrated with the machineries of the finishing line, that is, a spraying cabin with a robotic carousel, and an electronic tintometer system. Suitable experimental tests according to the Design of Experiments (DoE) are firstly defined, executed and analyzed for a series of color tones of interest. In order to derive and validate a set of colorimetric models able to evaluate and predict the color rendering of painted leathers, a set of recipes of basic dye pigments and data of light reflection measured by a multispectral camera are used. Principal Component Analysis is applied for dimensionality reduction, and linear least squares regression is employed to identify these data-driven models, which are then used for color control purpose. A color correction feedback strategy is indeed developed in order to converge towards the various target formulations. The control algorithm aims at reaching the multispectral reading values of the reference, that is, the first sample of unknown color recipe starting from the most similar archive base and appropriately updating the recipe of pigments, by using the measurement of leather samples prepared from time to time by the finishing line machineries. A set of company data are used to successfully validate the identified colorimetric models and the proposed color correction strategy.