AVR Virtual Environment in the New Generation Human Computer Interaction System

VE (Virtual Environment) provides the three dimensional visual interactive environment for the new generation the fourth generation natural HCI (Human Computer Interaction) system. Modeling the virtual objects is the key to build VE. The paper focuses on the theory, algorithms and core techniques for AVR (From Actual Reality to Virtual Reality) modeling of virtual environment. Based on the integration of computer vision and computer graphics (CV/CG), the objects in AVR VE are virtualized directly from the shape data (2D images and 3D data) acquired of their counterpart objects in real world by using AVR reconstruction approaches, which provides the AVR VE with a higher-level realism, conciseness and preciseness. Key-Words: Virtual reality, Virtual environment modeling, Superquadrics, 3D reconstruction, Human Computer Interaction 1 The new generation the fourth generation natural HCI system Human Computer Interaction (HCI) is the communication and interaction between human and computers. With the spread of computers and the rapid development of information technology, the network trends of information storage, the international exchange of information, and the diversity and individuality of individual information lead to an urgent need of natural, harmonious and simple human computer interaction similar to human beings. In the fourth generation HCI system [1,2], the traditional two dimensional interaction (text, character and image) of visual information develops towards the three dimensional visual interaction in a more realistic and immersed manner. As the key supporting technology, virtual environment (VE) will be the important interactive way in the fourth generation HCI, which provides three dimensional realistic visual interactive environment for HCI. Combining AVR research and HCI VE reconstruction, the paper focuses on the AVR theory, algorithms and technology in VE reconstruction, including AVR problem definition, application system framework and the key methods and implementation technology. 2 AVR and VE Modeling 2.1 AVR theoretical framework How to represent and understand the real world by computer is a general problem in many scientific and technological areas, especially in simulation, visualization, machine vision and intelligent information processing. Obviously, this is a type of consistent problem in nature, although in different areas, different attention has been paid to the problem, there is no systemic analysis and description for this problem. It is very necessary to systemically analyze the problem, build special technologic system and further study its theory and applications, Professor Yuan Baozong put forward a novel definition for this problem—AVR (From Actual Reality to Virtual Reality) problem [3,4], i.e. the problem of conversion from real world to virtual reality represented by computer, and the conversion techniques are named as AVR techniques. For building a common basis of the research on AVR theory and techniques, it is necessary to describe AVR problem in detail. Definition 1: AVR problem— the problem of conversion from the objects and surroundings in real world to the corresponding objects and surroundings in virtual world, which is described by the following form: ) ' , ' ( ) , ( i i T i i S O VR S O AR ⎯ ⎯ → ⎯ (1) where AR is Actual Reality, including i O (Objects) Proceedings of the 7th WSEAS International Conference on Simulation, Modelling and Optimization, Beijing, China, September 15-17, 2007 479