Abstract
To accelerate transformation in the clothing industry under Industry 5.0, in this study, a digital human model of regional young men was investigated, using a linear regression model approach. A total of 263 young men, supported by a 3D body scanning system, were selected for the experimental sample. Body indicators guiding a reverse model of the upper body were chosen, with body features identified through descriptive analysis and principal component analysis. Subsequently, a personalized linear regression model could be statistically analyzed, and personalized inverse models could be visualized by designing the algorithm flow of the digital human modeling system in Grasshopper. The model can be changed in real time to fit the concept of the human digital twin. Meanwhile, hierarchical clustering and fast clustering methods were used in combination to establish representative body shapes for young men in the region, and clustering centers were used as validation samples to demonstrate that the prediction accuracy of this digital human model system was more than 80% for most body indicators, validating the feasibility and reliability of the digital human model, based on linear regression. This study offers technical support for regional digital human modeling, establishes a model foundation for research on human surface details, and supplies a data reference for tailored clothing patterns.
Get full access to this article
View all access options for this article.