Skip to Main Content
In the past decade, fashion industry and apparel manufacturing have been applying intelligent CAD technologies to operate garment panel shapes in digital form. As garment panels are being accumulated gradually, there is a growing interest in finding similar panel shapes from large collections. The retrieved panel shapes can provide recommendations for stylists to reference and re-create during the cognitive fashion design process. In this paper, we propose a novel graph modeling and matching method to facilitate the searching of panel shapes for sketch-based garment design. A panel shape is first decomposed into a sequence of connected segments and represented by the proposed bi-segment graph (BSG) model to encode its intrinsic features. A new matching metric based on weighted direct product graph and minimal spanning tree (WDPG-MST) is then proposed to compute the similarity between two BSG models of the panel shapes. Finally in the front-tier, we provide a sketching interface based on our previous work for designers to input and edit the clothing panels. The simulation of the resulting garment design is also visualized and returned to the user in 3D. Experiment results show the effectiveness and efficiency of the proposed method.