Rendering

Overview

Rendering deals with creating images of objects as they appear in the real world, often with a strong emphasis on realism or interactive display algorithms. This is in contrast to Visualization, where the focus lies on the graphical representation of abstract or processed data, to enhance the understanding of situations that cannot be directly observed. Examples of rendering applications are virtual prototyping, architectural planning and reconstruction.
The rendering research group at VRVis addresses many problems associated with state of the art rendering tasks:

Acquisition – With methods like laser range scanning or photographic reconstruction, it has become feasible to capture very large areas with amazing detail. VRVis has significant experience with these processes and the associated challenges, including error detection, merging of partial acquisitions, data classification and simplification.

Scalability – Modern rendering systems are expected to display very large scenes with ease. These models are often in excess of several hundred megabytes, necessitating advanced resource management and highly efficient algorithms. The rendering system developed at VRVis is capable of displaying extremely large models – such as the entire city of New York – including highly detailed textures in real time.

Semantics – In addition to just displaying a model “as is”, it is often desirable to add some semantic information to identify or highlight specific parts for Visual Analysis. Handling semantic information has been an integral part of the rendering system at VRVis throughout its design and development, making it easy to adapt it to the specific requirements of new applications.

Interaction – Many applications do not simply provide a static scene, but require various levels of interaction. This could be a simple switch between several scenarios for an easy, direct comparison of design alternatives, or actually interacting with the model itself at any level, from repositioning parts to actually modifying its geometry and surface parameters. The highly flexible data structures at the basis of the VRVis rendering system allow its adaptation to these demands without any impact on its rendering performance.