Texture Modelling of Fabric Appearance Evaluation Based on Image Analysis

*Institute of Textile and Clothing, The Hong Kong Polytechnic University, Hong Kong Abstract This paper presents the development of a fairly new texture modelling method based on image analysis for the purpose of objective quality evaluation of fabric appearance. co-occurrence matrix analysis is used to give quantitative descriptions of fabric appearance properties; an expert system using a neural network is used to model the relationship between these essential features and the final rating grade of the fabric appearance. The experimental results demonstrate that a good correlation can be achieved between the actual rating grade and that predicted and reveals the possibility of the development of an artificial intelligence system to simulate the functions of the human eye and brain.