China
Africa
Explore chapters and articles related to this topic
Body and Senses
Published in David Burden, Maggi Savin-Baden, Virtual Humans, 2019
David Burden, Maggi Savin-Baden
Aliaga conducted a set of experiments to explore which factors contributed to the perception of cloth and to determine how efficiency of the modelling could be improved without sacrificing the perceived realism. Aliaga found that ‘appearance dominates over dynamics, except for the few cases where dynamics are very characteristic, such as in the case of silk’. (Aliaga et al., 2015). In addition to the demands from games and film CGI driving clothing modelling research, there is also increasing interest in cloth modelling in the fashion industry for both computer-aided garment design and to allow shoppers to try on fashion items virtually at home (called ‘Virtual Try On’ – VTO) before buying (Zhang et al., 2015).
The design of a remote fitting system for garment e-commerce
Published in The Journal of The Textile Institute, 2019
Zengrong Guo, Dongliang Zhang, Jituo Li, Siyuan Lin, Sirui Feng, Yi Xiao
With the development of 3D graphics and virtual reality, many companies and research institutions adapt virtual technologies to achieve virtual fitting. Virtual fitting technologies all need modelling and rendering of garment or human bodies. Many scholars put forward a series of human modelling methods. For example, Chen et al. (2016) improved the depth quality by registering the depth images captures from multi-views with a single Kinect and augment the human body information by fitting the statistical human model to the registered depth image. But in this way, a Kinect device must be equipped. Chu, Tsai, Wang, and Kwok (2010) proposed an exemplar-based method to generate 3D human body shape from semantic parameters. But the characteristic parameters are usually measured manually, which requires professional skills. Kasap and Magnenat-Thalmann (2011) proposed a method to obtain human models by deforming a human body template. However, the modelling results are deeply influenced by the configuration of the template model. Izadi et al. (2011) suggested KinectFusion based on a graphics processing unit, which is able to reconstruct static scenes with a real-time moving Kinect using implicit fusions and registrations. Zollhöfer, Martinek, Greiner, Stamminger, and Süßmuth (2011) presented a new method by means of establishing correspondences between depth and colour images of human faces, which can acquire high-quality human face models. But these modelling methods still take a lot of time to build consumers’ models, which leads to the high cost and long cycle. The variety, complexity and irregularity of fabrics make it difficult to model and simulate the cloth. Since the technologies of 3D garment modelling, physical simulation and rendering are limited, some researchers gradually put forward a series of cloth modelling methods and simulation methods (Chen, Feng, & Wang, 2013; Choi & Ko, 2002; Goldenthal, Harmon, Fattal, Bcrcovier, & Grinspun, 2007; Lee, Yoon, Oh, Kim, & Choi, 2010; Zhou, Jin, & Wang, 2010). Some companies and research institutions produced several kinds of products for virtual fitting, such as Metail’s virtual fitting website (http://www.metail.com). Although the simulation of the existing virtual fitting products looks good, there is a big gap between the simulation results and the real fabric. These unreal effects cannot convince the consumers. So, it is still very difficult for 3D virtual fitting to become the mainstream of fitting methods now.