A Method for Sharing Interactive Deformations in Collaborative 3D Modeling

Hiroaki Nishino, Atsunori SAKAMOTO, Kouichi Utsumiya, Kazuyuki Yoshida

Department of Computer Science and Intelligent Systems,
Oita University,
Japan

Kazuyoshi Korida
Department of Communication,
Oita Prefectural College of Arts and Culture,
Japan

A distributed virtual environment (DVE) is a potential technology to bring out a fundamental change in traditional design and production methodologies. The traditional methods used to divide a whole process into multiple stages and only skilled persons serially carried out each stage. The introduction of the DVE technology, however, enables various people such as designers and developers in different fields, external consultants and even customers to be involved in collaborative design and production activities. There are several frameworks proposed to make the DVE be usable and practical on the network such as FIVE[1], NPSNET[2] and Shastra[3]. There are, however, few trials to applying the DVE technology especially for early conceptual design of new products such as the creation of unprecedented shapes and patterns.

This paper proposes a new approach to constructing a multi-user collaborative environment for conceptualizing and designing new products and crafted objects. To support creative activities in the multi-user environment, we especially pay attention to Japanese ceramics, a typical handicraft, as a target application of the collaborative work, proposing a concept, the so-called skill sharing. In the traditional art and craft fields, there are some tendencies to keep traditional ways being followed for hundreds of years and protect their own styles from others. It is, however, quite useful to share not only digitized art works, but also their cultivated skills, experiences and sixth sense acquired from long-term efforts and disciplines among all participants involved in a collaboration group. The sharing of such undocumentable knowledge requires to visualize a series of the expert's modeling operations including which primitive shapes to use, and how combine, blend and deform them to make up a new object rather than just sharing a final form. Because many shapes and patterns are created or modified one after another in the course of the modeling process, these intermediate shapes need to be shared efficiently. On the other hand, a standard 3D data format available on the Internet such as VRML tends to be large in size, making the sharing of rapidly deforming 3D objects a difficult task. Therefore, we extended our intuitive 3D modeling environment[4] to be usable on the network by adding the functions to allow multiple participants working on the same objects and sharing all intermediate information during a collaboration session in real time. The system facilitates the designers to intuitively specify shapes and deformation patterns by using their bimanual gestures, providing a 3D modeler to generate their desired shapes according to the gesture input.

All shapes generated in the course of the modeling session are represented as implicit surfaces of blended primitives. The primitives are defined as superquadric functions extended to incorporate some essential deformation capabilities such as pinching, tapering, twisting and bending. Because this 3D object representation method can express various shapes by using a small number of functional parameters (usually dozens of them), only these parameters need to be shared among the client systems on the network. If some modeling operations performed on a participating client, then only modified parameter values need to be updated. It is a client's responsibility to convert the set of parameter values to a ready-to-visualize data format such as the polygon mesh. Then, each client can generate the 3D models in the appropriate resolution according to its CPU and graphics power. This mechanism offloads some server's burden like the network bandwidth monitoring and the shared data resolution (LOD) control usually required on the related DVE systems. It puts the clients to be more practical use as well as making the system be scalable. It also enables the geographically dispersed people to participate in the collaborative session through wireless links as if they work together face-to-facedly in a distributed virtual environment.

The final paper describes the functional requirements and an implementation method of the proposed approach.

Reference

[1] Slater,M. and Wilbur,S.: A framework for immersive virtual environments (FIVE): Speculations on the role of presence in virtual environments, Presence, 6(6):603-616, 1997.

[2] Macedonia,M.R. et al.: NPSNET: A network software architecture for large scale virtual environments, Presence, 3(4):265-287, 1994.

[3] Anupam,V. and Bajaj,C.L. : Shastra: Multimedia collaborative design environment, IEEE Multimedia, 1(2):39-49, 1994.

[4] Nishino,H., Utsumiya,K., and Korida,K. : 3D object modeling using spatial and pictographic gestures, Proc. of the ACM VRST 1998, pp.51-58, 1998.