From body measurements to body shape perception: an intelligent tool for garment design

This paper presents a model characterising the relation between emotional descriptive keywords describing human body shapes and concrete body measurements. Three procedures have been proposed for building this model. In the first procedure, the experts generate a list of emotional descriptors describing human body shapes and then evaluate a set of virtual human bodies of different shapes using these descriptors. Next, two algorithms of decision tree (CART and fuzzy-ID3) have been applied for modelling the relation between each emotional descriptor and the ratios of relevant body measurements. In the second procedure, the experts evaluate the relationship between these concrete emotional descriptors and a number of abstract fashion themes such as ‘sporty’ and ‘attractive’ without taking into account any specific human body shape. This conceptual relationship given by different experts is modelled using a fuzzy cognitive map. In the third procedure, the two previous models are combined using the fuzzy relation operations. Using this combined model, garment designers can quickly identify personalised and variable body shapes, expressed by a set of concrete and abstract keywords, from human body measurements. These keywords will be further integrated into an expert knowledge base of garment design for developing personalised new products.