avatars is a strenuous activity which requires hours
of work even for the experienced 3d artist. This is a
limitation in the broad adoption of H-Anim
specification. Furthermore, there is a significant
number of NVE communities that are not providing
a large number of available avatars to their potential
users. For example open source or research projects
cannot afford to hire 3d designers for this task. This
discourages users from participating to such a
community since users want to project their
individuality through their avatar in a virtual world.
A convenient solution to this problem is to provide
one type of avatar that can be customized in a
graphical user interface by the users according to the
his/her preferences. This solution not only saves
resources, but it also allows users differentiate their
avatars.
This paper is structured as follows. The next
section presents similar tools. Next we are
describing the main design characteristics of
HanimE and the functionality it offers. Afterwards,
we are presenting implementation and
interoperability issues and the way we have faced
them. Finally, some concluding remarks and planned
next steps are briefly described.
2 RELATED WORK
Most of the tools for visual H-Anim avatar editing
are either focused to specific tasks or complicated
for the average user. On the one hand this restricts
their integration to NVEs. On the other hand, it
discourages the average user from using them. Wang
and Ressler (2007) present a tool for viewing and
manipulating CAESAR (Civilian American and
European Surface Anthropometry Resource)
generated bodies. However, this work does not apply
to the average user neither from the perspective of
functionality nor from the perspective of ease of use.
Cobo and Biery (2002) presented a Web3D
toolbox for creating H-Anim compatible actors. The
toolbox supports avatar modelling through
primitives, meshes and NURBS, joint and facial
animation. It relies on Shout3D, which is a
proprietary and commercial platform for the
visualization of the avatars. This limits its portability
and ease of integration to NVEs. In addition it
requires that the user has 3D modeling skills and
intermediate knowledge of the H-Anim standard, in
order to use it effectively. HanimE editor presented
in this paper differs in two ways. First of all, it is
based on non-proprietary plug-ins, which allows
easy integration to third party platforms. Secondly, it
is crafted in a way that the basic avatar processing
tasks are performed in a way that no H-Anim
specific knowledge is required from the user.
3 MOTIVATION AND
RATIONALE
Our solution aims at bridging the gap between the
user’s demands for simple avatars creation and the
compatibility with H-anim standard. For that reason
we have implemented HanimE, that is offering to the
avatar creators a graphical interface to create their
avatars. All the editing takes place on a purely visual
level, providing the user with all the sufficient
graphical tools required so that s/he can see any
change on his avatar instantly. Possible changes
include the change of the color of avatar’s shoes, the
exact fitting of sun glasses on the avatar’s head, or
the complete replacement of a body part with a new
one. Any coding and/or technical issue concerning
H-anim compatibility is transparent to the users.
HanimE supports the following features:
• Automatic or manual H-Anim avatar loading.
• Manipulation of the Material Node properties
such as colour, transparency etc.
• Automatic or manual H-Anim avatar scaling.
• H-Anim avatar rotation.
• Texture mapping.
• Virtual items library.
• Adding and adjusting size and orientation of
virtual items on an avatar.
• Save X3D Scene to file.
HanimE is organized in three parts: (a) a
visualizer; (b) library containing virtual items/ H-
Anim body parts; and (c) a material node editor.
This organization aims in guiding the user to the
desired result through simple steps that are easy to
follow.
The visualizer is the first user interface screen
that appears when H-AnimE is launched. An avatar
can be loaded both automatically or manually.
Finally, from that screen a user can save his/her
avatar. The visualizer is an extended VRML/X3D
browser window that deals with the tasks of loading
an avatar, visualizing it and exporting it to a file. In
addition this window triggers the virtual items/ H-
anim body parts library window and/or the material
node editor. The loading portion is accomplished by
a custom fully featured X3D/VRML parser. This
parser enumerates various aspects of an VRML/X3D
file such as characters, lines H-Anim Joints,
Segments Humanoids, Displaces and Sites.
GRAPP 2010 - International Conference on Computer Graphics Theory and Applications
348