Local Random Phase Noise
For Procedural Texturing
G.Gilet & B. Sauvage & K. Vanhoey
& J-M. Dischler & D. Ghazanfarpour

Your browser doesn't support the features required by impress.js, so you are presented with a simplified version of this presentation.

For the best experience please use the latest Chrome, Safari or Firefox browser.

Local Random Phase Noise

For Procedural Texturing

G. Gilet & B. Sauvage & K. Vanhoey & J-M. Dischler & D. Ghazanfarpour

Context

Increasing needs for large and detailed scenes
Manual creation is painstaking
Need automatic content creation
By example texturing

Texture synthesis

Noise-based procedural texturing

Pros

No Repetition
Compact
Continuity
Generic
Fast

Noise-based procedural texturing

Cons

Hard to control
- By-example
Does not reproduce structure

Our key contributions

Fast evaluation
Reproduce some structure
Automatically generated by example

Noise By Example Fundamentals

Noise By Example Fundamentals

Noise By Example Fundamentals

Gabor Noise By Example [Lagae et.al 2012]

Let's reproduce some structure

Structure is encoded in the phase
We need to preserve some phase information
Not possible with Gabor Noise By Example

Noise Formulation : the details

IFT

Noise Formulation : the details

IFT

Noise Formulation : the details

IFT

IFT

Tradeoff speed / spectral accuracy
Reproduce structure

Reproducing Structure

Results : regular textures

reg

Comparison with patch based

reg

Comparison with patch based

reg

Texturing 3D model

Pong

180 fps

Texturing 3D model

Texturing 3D model

Pong

90 fps

Texturing 3D model

Pong

10-40 fps with tessellation

Limitations

Not strictly stationary
Parameterization dependent
Structures spread all over the spectrum

Conclusion

Reproduction of structured patterns
By example
Efficiency

Future works

Better tools to identify the structure
Extension to 3D

Thanks for your attention

Local Random Phase Noise