为实现青年女性乳房形态的有效判别,提出了以横、纵截面角度指标细分青年女性乳房形态的方法.使用三维人体扫描仪获取209名青年女性胸部点云数据,提取表征乳房挺拔度、丰满度及外扩度的3个角度参数及其他相关参数;采用主成分分析降维,得到影响乳房形态的4个特征因子,用混合F统计量确定乳房形态聚类的最佳分类为5类.根据分类结果构建乳房形态的Fisher判别模型,整体判别精度为96.23%;在青年女性乳房中,占比较大的依次为饱满挺拔型(34.9%)和适中型(22.5%),平坦外扩型占比最小,为6.7%.该研究结果可为青年女性内衣等产品设计提供参考.
To effectively discriminate the breast morphology in young women,a method of subdividing breast morphology in young women based on cross-sectional angle indicators was proposed.A 3D human scanner was used to obtain point cloud data of 209 young women's breasts,extracting three angle parameters and other related parameters that characterize breast height,fullness,and expansion.The principal component analysis was used to reduce the dimension,and four characte-ristic factors affecting the breast shape were obtained.The mixed F statistics was used to determine that the five categories of breast shape clustering were the best.According to the classification results,a Fisher discriminative model for breast shape was constructed,and the overall discriminant accuracy was 96.23%.Among young women's breasts,the larger proportion is 34.9%for the plump and straight type,and 22.5%for the moderate type,with the smallest proportion being the flat outward expansion type,which is 6.7%.The research results can provide reference for the design of young women's underwear and other products.