Application of Cluster Algorithm in Clothes Shape Classifying
Xiao Gang Wang, Xin Zhan Li, Yue Li · Applied Mechanics and Materials · 2011
Based on research about patterns of garment, patterns were made to achieve data and interval of the bust eases. On the basis of bust eases, a series of shape profiles in different eases of garment were designed and distinguishing experiment was done according to the theory of psychics, which profiles in different eases were distinguished. The shapes in different fit were classified into four clusters: tight fit, fit, little loose and loose. The results of experiment were analyzed by k-means cluster method and quantitative classification based on bust ease was achieved. It opens our mind to make garment research by data mining method. A new method for garment fit research and classification exploration of outline shape was brought forward, which it offers reference for garment industry and research for automatic computer distinguishing technology.