METHOD OF DETERMINING TEXTURE PARAMETERS OF LACQUER Russian patent published in 2019 - IPC G01J3/46 G01N21/29 

FIELD: paint industry.

SUBSTANCE: invention relates to a method of predicting visual parameters of a lacquer texture with a known lacquer formulation, as well as a corresponding computer program for carrying out the disclosed method on a computing unit. Disclosed is a method of predicting visual parameters (9) of a lacquer texture with a known lacquer formulation (1), wherein based on a number of paint-forming components used in lacquer formulation (1), an artificial neural network (7) determines visual parameters (9) of the lacquer texture. For the known lacquer formula (1), physical model (3) is used to determine the value of at least one characteristic value (5) describing at least one optical property selected from the following list of optical properties: spectral reflection of lacquer formed based on the corresponding lacquer formulation, an optical constant of at least one coupler-forming component of the corresponding lacquer formulation, predicted by physical model reflection spectrum of at least one coupler-forming component of corresponding lacquer formulation, Kubelka-Munch absorption component K, Kubelka-Munk scattering component S and at least one optical characteristic value calculated from the corresponding optical constants of the several coupler-forming components of the corresponding lacquer formulation. Value of at least one characteristic value (5) is correlated with the known lacquer formula (1) and transmitted as an input signal of an artificial neural network (7) for determining visual parameters (9) of the texture. Value defined and associated with the known lacquer formulation describes at least one optical property for at least part of the lacquer forming components of lacquer formula (1). For training neural network (7) using samples of paints with respectively known lacquer formulation, wherein for each of the ink samples, its corresponding visual texture parameters are measured and correlated with the value of at least one characteristic value, which is defined for the corresponding lacquer formulation, which describes at least one optical property for that corresponding lacquer formulation.

EFFECT: high accuracy of predicting lacquer texture parameters.

10 cl, 2 dwg