Material Texture Image Model Optimization and Result Analysis

Abstract

Targeted in the quantitative description of the relationship between the materials subjective features and objective parameters, this study builds a mathematical model by BP neural network. Then optimization of the thresholds and weights of BP material texture models is conducted to refine the accuracy and description ability of this network. Through the analysis of the result of GA-BP Model, the foundation established by the summary of relationship between the texture image and the objective material parameters can be used to forecast the emotional characters of the materials whose objective parameters are previously understood.