A Study on Analyze Body Characteristics and Predict Human Size of Young Women from Northeastern China

The MTM manufacture mode is becoming more and more widely used in the clothing enterprise. How to obtain efficient and accurate data about the human body becomes the most important step in the MTM mode. In this paper, the body characteristics of young women aged 18–25 years old in Northeast China were studied, which used the non-contact 3D body measurement technology to get the date for 450 young women’s body characteristics. The results show that their average height is 162.15 cm, and 15 characteristic values of the other body parts are given. Select 370 groups as training data, 80 groups as test data, and use the linear model and the neural network to predict the body characteristics on the basis of the incomplete data. To be more specific, based on the analysis of human body characteristics and the method of body type classification, select the data of 5 key body parts from 15 human features (height, weight, chest circumference, waist circumference, hip circumference) as input data and 10 parts data as output data, use the regression equation listed by linear regression and neural network algorithm to establish the human body data forecasting model to predict the output value, and then validate and optimize the result of the experiment. Experimental results show that the prediction error to the human body of the prediction mode is within the scope of the permit. The contrast test shows that the method is effective to predict human body data for clothing purpose, and with the increase of the training sample amount, forecasting accuracy is significantly improved.