Virtual technology of Made-to-Measure Men Shirt

Abstract

The aim of this research is to propose the enhanced method of pattern block drafting with novel pattern-oriented body measurements and to optimize the fit of men shirts and customization process that demonstrating by the virtual system “body-garment”. To investigate and devise this method, 119 young male bodies were scanned by 3D body scanner. New anthropometric indexes were proposed according to the theories and basic rules of sketching patterns[1]. Competitive neural networks and self-organized mapping neural networks2(SOMNN) were trained and tested to clustering the male body features into certain types. Thereby novel patterns for diverse types were employed as the database of customized men shirts. Compared with the traditional pattern blocks, the new ones are providing the better fit in virtual environment. Based on the proposed method, the individual pattern could be generated immediately from the existed database with facilitated precision and efficiency, by matching the body measurements of customers to certain anthropometric features.