A novel optimization approach to minimize aggregate-fit-loss for improved breast sizing

Abstract

Ready-to-wear clothing is typically based on the body-shape of human fit models that an apparel company hires. The body-shape difference between a consumer and the fit model of their size results in fit-loss of a certain degree. Aggregate-fit-loss is a concept attempting to quantify and estimate the accumulative fit-loss that a population may encounter. This paper reports on a novel method that minimizes the aggregate-fit-loss of a sizing system for bras, through shape categorization and optimized selection of prototypes (which can be regarded as the most appropriate fit models, or standard dress forms) for the categorized groups. A fit-loss function was introduced that calculates the dissimilarity between any two three-dimensional body scans, via pointwise comparisons of the point-to-origin distances of 9000 points on the scan surface. The within-group aggregate-fit-loss is minimized by an algorithm that returns the optimal prototype for the group. The overall aggregate-fit-loss is reduced by breast shape categorization based on the dissimilarities between the scans. Finally, the constraint of band sizes was brought into the categorization to provide a more feasible solution for improved bra sizing. The findings of this study can also contribute to the optimization of sizing systems for other apparel products.