A study for image-based integrated virtual environment

In this paper, we propose a new approach to building a virtual environment by using raw data from a real environment. To manage and analyze different types of datasets from many sensors and databases, we build and integrate different types of environments with different types of datasets, user interactions and rendering methods. To integrate the environments, we merge rendered images of each virtual environment according to the user´s viewpoint, depth information, and reliability of rendered images, and display the resultant image to the user. In addition, we propose an integration environment by using an environment manager, which determines the user´s viewpoint with respect to the integrated environments according to the user´s input, and compose a display image from rendered images of the integrant environments. The reliability of the rendered images is determined by data accuracy and resolution, display resolution and rendering errors from rendering methodologies. By integrating four constructed environments according to the user´s viewpoint, we can construct a photorealistic large-scale environment from different types of datasets. Furthermore, we implemented the prototype system in CABIN and demonstrated the scalability of this architecture.