Object based manipulation with 3D scenes in mobile environment

Abstract

The increasing power and display resolution of mobile devices allow the user nowadays to work with 3D information in mobile environment. The use of this new technology brings some new problems that need an urgent solution. One of them has its roots in the fact that common users are not trained to work in 3D graphical environment in general. The main obstacle for a common user is the fact that 3D environment offers too much freedom for object manipulation in comparison with situation in 2D environment. 

There are various solutions to this problem but usually they do not handle efficiently the problem of navigation in 3D environment in context of handling appropriate information density on a small screen on a mobile device. Our approach is based on transformation of a 3D scene in 2D representation in order to decrease freedom of movement during navigation in the scene. The navigation (and other types of interaction) is performed in 2D environment - the information acquired during interaction is transformed back into 3D form (into 3D object space). In such a way the user can manipulate with single objects in various modes (e.g. the user can change focus on various objects what might e.g. result in their temporary elimination from the scene).

The work in 2D in mobile environment requires some special features that are not considered in standard environment. Especially critical is zooming. Zooming in raster form could change the picture on a small screen into completely unusable representation (pixels will be transformed in tiles that will be difficult to perceive). Solution to this (and to other related problems) is the use of SVG format to handle 2D scene representation. The main advantage of SVG is that it is object oriented and supports zooming, interaction with objects, scene annotation and manipulations with objects.  The process of scene transformation and manipulation with objects is distributed between the server and mobile device. The transformation is performed on the server side. The resulting 2D representation of the scene is sent to mobile device where the visualization is performed. The user can interact with the picture on the mobile device with interaction techniques available on mobile devices.

As the SVG is XML based format, the Cascading Style Sheets can be used to style all graphical elements. Therefore each object can be visualized in different mode (solid, transparent, semitransparent). These modes could be changed interactively. This feature (interactive change of modes) substitute in certain sense some functions available for avatars during their walkthroughs when performed in standard 3D environment (e.g. "having look behind corner" etc.). The user interaction in 2D environment is rather limited in comparison with standard 3D interaction but it is less demanding for the user. On the other hand it brings better picture quality due to SVG features applied e.g. during the process of zooming.