Abstract |
The production of different Motion Visualizations in 2D or 3D is a topic that has gained increasing
attention by the graphics, design, animation and gaming industry. Until now, interaction of Virtual
Humans (VHs) with tools and machines has been mostly addressed through predefined
animations of both the object and the VH, or through the usage of physics simulation engines.
Recently, such visualizations have gained increasing interest in the domain of Digital Cultural
Heritage (CH). In this setting, Motion Visualization and Motion Capture technologies have come
to the spotlight, mainly through the need to capture intangible dimensions of CH existing in
theatrical performances, dances and crafts. Despite the cultural, historical, economical and
societal significance and value of Heritage Crafts, efforts towards their digital representation,
presentation, and preservation are scattered. To that end, this Thesis proposes an approach
towards visualizing craft processes in Virtual Environments (VEs). In the proposed approach the
practitioner is represented by a VH and objects through their 3D reconstructions. Craftsmen’s
actions are reproduced by animating the VH based on Motion Capture (MoCap) recordings, while
the motion of machines and tools is induced by the human motion. The appropriate simulation of
VHs is an important aspect, since crafts are practiced by humans and machines are designed for
use by them. At the center of the proposed approach is a conceptual, twofold decomposition of
craft processes into actions, and of the machines used into components; this is essential in the
systematic transfer of craft practice from the physical to the virtual domain, while retaining
realism and allowing the semantic representation of craft processes. As an application of this
approach, this thesis also presents the MoViz platform, allowing users to create and experience
craft usage scenarios. In particular, MoViz enables users to author their own scenes, where
actions and machine parts are assigned to VHs, in order to create Motion Vocabularies,
representing the craft process. Users can also experience playback of the created scenes, through
simulation in a VE, either in 3D or in Virtual Reality, where a training mode is also available.
Through the aforementioned process, we aim to deliver an efficient way of visualizing craft
processes within VEs, thus increasing the usability and educational value of craft representation,
and opening the way to a variety of new applications for craft presentation, education and
thematic tourism, based on the value of tradition and intangible cultural heritage. Moreover, this
work aims to contribute to the preservation and perpetuity of not only the craft, but also of the
legacy of the practitioners, whose unique movements are preserved via the MoCap. In the context
of this Thesis the main focus is on the craft of loom weaving; however, the proposed methodology
is generic, so that the MoViz platform can be used for the representation of any craft.
|