Men's Shirt Pattern Design: Part II: Prediction of Pattern Parameters from 3D Body Measurements

Abstract

Part I of this two part series of papers has shown that existing pattern drafting methods are much less than adequate for drafting patterns to fit a wide range of body morphology. To solve this problem, this paper considers predicting shirt pattern parameters from 3D body measurements. Two prediction models are reported in the paper. One is established using multiple linear regression and the other using the Artificial Neural Network (ANN). It shows that the ANN model can predict the pattern parameters very accurately, but the linear regression model has the advantage of showing the relationship between the pattern parameters and the specific body measurements. This work is believed to be important to the implementation of apparel mass customization.