Generation of Body-Fit Garment Patterns Using a Landmark Matching Algorithm

Advancements in 3D technology have largely automated the body measurement process for garment production. However, drafting body-fitting patterns still depends on the intuition of the pattern designers. In this study, an alternative method has been developed that generates garment patterns automatically by analyzing the relationship between the garment and the underlying body model. For this, a shortest-path finding algorithm was developed to extract corresponding measurements from the garment and body. Then the garment patterns were defined in a polar coordinate system using the measured data. Finally, the similarity between template and target garments was analyzed using an image-based method. This method is expected to facilitate the generation of personalized garment patterns, offering insights into the application of virtual and 3D digital technologies in fashion production.