公开(公告)号：US20190078936A1

公开(公告)日：2019-03-14

申请号：US15771577

申请日：2016-10-27

申请人： BASF Coatings GmbH

发明人： Guido Bischoff ,  Martin Schmitz ,  Donald R Baughman

IPC分类号： G01J3/46 ,  G01J3/50 ,  G06N3/08

CPC分类号： G01J3/463 ,  G01J3/46 ,  G01J3/462 ,  G01J3/504 ,  G01N2201/1296 ,  G06N3/084

摘要： Described herein is a method for predicting visual texture parameters of a paint having a known paint formulation. The visual texture parameters of the paint are determined using an artificial neural network on the basis of a number of color components used in the known paint formulation. The method includes determining a value of at least one characteristic variable describing at least one optical property using a physical model for the known paint formulation. The method also includes assigning the value to the known paint formulation, and transmitting the value to the artificial neural network as an input signal for determining the visual texture parameters. The value describes the at least one optical property for at least some of the number of color components of the known paint formulation. The method further includes training the neural network using a plurality of color originals each having a respective known paint formulation.

公开(公告)号：US10401224B2

公开(公告)日：2019-09-03

申请号：US15771577

申请日：2016-10-27

申请人： BASF Coatings GmbH

发明人： Guido Bischoff ,  Martin Schmitz ,  Donald R Baughman

IPC分类号： G01J3/46 ,  G06N3/02 ,  G01J3/50 ,  G06N3/08

摘要： Described herein is a method for predicting visual texture parameters of a paint having a known paint formulation. The visual texture parameters of the paint are determined using an artificial neural network on the basis of a number of color components used in the known paint formulation. The method includes determining a value of at least one characteristic variable describing at least one optical property using a physical model for the known paint formulation. The method also includes assigning the value to the known paint formulation, and transmitting the value to the artificial neural network as an input signal for determining the visual texture parameters. The value describes the at least one optical property for at least some of the number of color components of the known paint formulation. The method further includes training the neural network using a plurality of color originals each having a respective known paint formulation.