Inria / Raweb 2004
Project-Team: EVASION

Search in Activity Report, year 2004:


Project-Team : evasion

Section: New Results

Efficient rendering of natural scenes

Participants: Florence Bertails, Marie-Paule Cani, Philippe Decaudin, Alexandre Meyer, Fabrice Neyret.

Volumetric representation using texcells

Participants: Philippe Decaudin, Fabrice Neyret.

In the scope of the Vertigo collaboration (RIAM funding, cf  7.1), we are extending our past work on realistic, real-time volumetric textures, the target being the management of forest covered scenery in virtual reality applications (see figure 9).

Figure 9. Real-time realistic forests using Volumetric textures (mixed together with Bionatics' hybrids).

We have developed a new representation allowing for the real-time rendering of realistic forests. It produces dense forests corresponding to continuous non-repetitive fields made of thousands of trees with full parallax. It draws on volumetric textures and aperiodic tiling: the forest consists of a set of edge compatible prisms containing forest samples which are aperiodically mapped onto the ground. The representation allows for quality rendering, thanks to appropriate 3D non-linear filtering. It relies on LODs and on a GPU-friendly structure to achieve real-time performance.

This resulted in a publication [15] at EGSR'04. Moreover, the Master thesis work of Florent Cohen on shading and shadowing the forest model with the GPU led to a poster [31] presented at Siggraph'04.

Concerning the Vertigo collaboration, a library combining our representation and the "hybrid" trees developed by Bionatics has been realised. It has been integrated in a prototype based on the professional training flight simulator developed by Thales Training and Simulation.

Adaptive surfels for real-time forest sceneries

Participants: Alexandre Meyer, Fabrice Neyret.

During his Master Thesis, Guillaume Gilet has developed an adaptive model of surfels (a point based representation): the size of points (i.e. discs) depends on the distance and on the visibility, and represent a set of leaves (these sets are organized hierarchically in that purpose). Moreover, surfels commute to classical meshes for close points of view. This allows for the interactive rendering of forests with both close and distant trees, and continuous flyovers of entire forests (see Figure 10).

Figure 10. Interactive forest rendering utilising surfels

Self shadowing of animated scenes

Participants: Florence Bertails, Marie-Paule Cani.

Self shadows are particularly important to get the adequate impression of volume for complex natural objects such as hair (see Figure 11). We just developed an efficient self shadowing method (submitted for publication) particularly well adapted to the rendering of animated objects, since it requires no geometry-based pre-computation. Our method is based on a 3D light-oriented density map, a novel structure that combines an optimized volumetric representation of hair with a light-oriented partition of space. Using this 3D map, accurate hair self-shadowing can be interactively processed (several frames per second for a full hairstyle) on a standard CPU. Beyond the fact that our application is independent of any graphics hardware (and thus portable), it can easily be parallelized for better performance. A parallel implementation makes the method run in real-time.

Figure 11. Hair rendering without (left) and with self shadows (right).