Cognitive Garment Panel Design Based on BSG Representation and Matching

Cognitive Garment Panel Design Based on BSG Representation and Matching

Shuang Liang, Rong-Hua Li, George Baciu
DOI: 10.4018/jssci.2012010104
(Individual Articles)
No Current Special Offers


Previously, the fashion industry and apparel manufacturing have been applying intelligent CAD technologies with sketching interfaces to operate garment panel shapes in digital form. The authors propose a novel bi-segment graph (BSG) representation and matching approach to facilitate the searching of panel shapes for sketch-based cognitive garment design and recommendation. First in the front-tier, they provide a sketching interface for designers to input and edit the clothing panels. A panel shape is then decomposed into a sequence of connected segments and represented by the proposed BSG model to encode its intrinsic features. A new matching metric based on minimal spanning tree is also proposed to compute the similarity between two BSG models. 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.
Article Preview

1. Introduction

For the last few decades, computer-aided design (CAD) systems for apparel manufacturing have been rapidly developed and have become the basis for textile garment production (Aldrich, 2007). Traditional CAD systems impose steep learning curves and complex interactions, often leading to multiple non-intuitive intermediate steps. Fashion designers are often required to practice unnatural command sequences and navigation paths in order to complete basic shape adjustments. Shapes are drawn by low-level point-and-click processes, not strokes. These systems lengthen the design cycles at a very fundamental level, as they have significantly departed from the natural free-hand sketching. Therefore, it is desired to facilitate the CAD systems with more intuitive, natural and efficient input manners, i.e., the sketching interfaces, to create apparel designs. Sketching is an artistic medium with its low overhead in designing, representing, and communicating graphical ideas. Sketching interfaces are favored in conceptual processes, especially for designers to present representations that are fundamentally different in spirit and purpose. Along with the fast development of pen-based and touch-based devices and techniques, garment sketching design attracts lots of attentions from designers, stylists, and researchers. However, due to the ambiguity and free-style of sketching inputs, it requires the system to recognize and understand the user’s input sketching shapes cognitively and effectively.

Meantime, a primary task in computer-aided garment design has been how to construct garment panels, i.e., strips of fabric in garment. These panels are usually produced by pattern technologists (tailors) according to the paper designs drawn by fashion stylists before they are sewn together to compose a complete dress. As the number and variety of the designed garment panels continue to grow, these panel shapes substantially form a knowledge-base for reference in future garment design. New garment designs can be made by re-editing or revising these existing shape designs. This is extremely useful and helpful in the fashion industry, where the popular element of dress designing not merely dates back to last century. Prevailed by the trend of back-to-ancients, applications to help people manage these panel databases are attracting increasing interest. In the near future, the challenge of garment design will shift from “How to design new garment panels?” to “How to manage and find the existing designs for further re-creation?” The goal to search for similar panel shapes from large collections is shared not only by designers, technologists and professionals, but also by general users.

In this paper, we propose a new cognitive garment panel design approach based on Bi-Segment Graph (BSG) representation and matching. Our goal is to return similar garment panels cognitively and intelligently according to the user's sketching input shape. There are three main techniques in our method: (1) First, in the front end, we provide a touch sketching interface for panel shape input and edit. (2) We then propose a new bi-segment graph (BSG) model to further formulate the bi-segment structure in our previous work (Liang, Chan, Baciu, & Li, 2010; Liang, Li, Baciu, Chan, & Zheng, 2010) into weighted graphs. The BSG model describes the panel shape by a sequence of adjacent segments and the encoding features among them, which can be calculated and matched by machine cognitively. (3) Finally, in order to perform the searching task, we propose a new matching metric based on minimal spanning tree to compute the similarity between two BSG models. We also extend our previous work (Wong & Baciu, 2005) to support 3D visual feedback for panel matching and garment simulation. Figure 1 illustrates the working scenario of garment panel design and matching process.

Figure 1.

Illustration of the working scenario for garment panel design and matching process. (a) A sketch input of the front panel shape for a skirt. (b) The panel matching result which shows the corresponding regularized skirt shape from our panel database. (c) The visual feedback and simulation of the skirt mapped onto a 3D body model.


Complete Article List

Search this Journal:
Volume 15: 1 Issue (2023)
Volume 14: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 13: 4 Issues (2021)
Volume 12: 4 Issues (2020)
Volume 11: 4 Issues (2019)
Volume 10: 4 Issues (2018)
Volume 9: 4 Issues (2017)
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing