A new shape clustered leg sizing system for mass customization fit of compression garments

Abstract

Effective identification of body shape and size is essential for designing fitted compression garments. However, existing leg sizing systems often neglect shape variations. This study developed a novel shape-clustered leg sizing (SCLS) system for mass customization fit of compression garments. Applying 3D digital body scanning technology, we analyzed the anthropometrical features of 480 lower limbs from 240 adults (mean age 55.16 ± 4.65 years). Key features, such as gradients and angles of turning points that define the outline curves of the lower limbs, were employed to classify leg shapes above the knee, at the knee, and below the knee, using the K-means method that was selected from nine different clustering algorithms based on clustering quality assessment. Size distributions for each leg shape were quantified using a multiple-percentile approach. The SCLS system identified 12 distinct leg shapes and 8 sizes per leg shape, highlighting morphological variations in the lower extremities. This approach provides a sophisticated and practical sizing method that enhances garment fit by inclusively considering both shapes and sizes. The developed leg shape-based sizing matrix also identified groups of highly correlated leg shapes, facilitating rapid dimensional transformation across shapes and thereby improving fitting efficiency in the mass customization of compression garments.