## Section: New Results

### Scientific visualization

Participants : Luc Buatois, Matthieu Chavent, Bruno Lévy, Nicolas Ray, Bruno Vallet.

#### Geometry textures

Last year, we introduced
*geometry textures* , a novel geometric representation for surfaces based on height maps. The visualization is done through a GPU ray casting algorithm applied to the whole object. At rendering time, the fine-scale details (mesostructures) are reconstructed preserving original quality.
Visualizing surfaces with geometry textures allows a natural LOD behavior. This year, we explored new applications of geometry textures, and published them in
[18] . We extend the application of geometry textures beyond their initial motivation (mesostructure visualization). Thus, we present three new applications: rendering of
solid models, geological surfaces visualization and surface smoothing.

#### Visualization of CAD and industrial models

We continued our work on the visualization of large-scale CAD models for industrial structures. In the specific case of the oil and gas industry, structures comprise a huge number of pipes. We developed techniques to recover the equations of cylinders, cones and toric sections from triangle soups, and a GPU-oriented ray-caster for high-quality and efficient rendering of these structures. We published the method in [21] . Some results are shown in Figure 8 .

#### Molecular visualization

We continued our work on interactive molecule visualization. Our system visualizes the so-called MSS (Molecular Skin Surface). Our approach is based on a GPU ray-caster, that directly uses the equation of the MSS (a piecewise quadric surface), rather than discretizing it into triangles (Figure 9 -left). This year, we improved the efficiency of the algorithm and published an article in the Journal of Molecular Graphics [17] . We also implemented an efficient version of a specific global illumination effects (screen-space ambient occlusion), as shown in Figure 9 -right. This was done in the frame of the Ph.D. of Mathieu Chavent and the training period of Cécile Poisot.