Study on Fabric Classification with Support Vector Machines Based on Half-Circle Skirts Shape Simulated in 3D Virtual Try-On

摘要

The article studies the shape of half-circle skirts made of 55 kinds of typical skirt fabric. 3D virtual try-on platform is applied to simulate the shape effect of skirts when they are worn and the samples are cluster analysed according to the shape index of half-circle skirts. 45 kinds of fabric are randomly selected to establish the classification model with Support Vector Machines (SVMs) based on the relationship between mechanical properties of fabric and category of shape style of half-circle skirt. Then the model is used to verify and evaluate the remaining 10 kinds of fabric. The results show that the model can preferably predict the fabric classification of half-circle skirts shape.