![](bg8.png)
tures with very different levels of intensity variation.
Color restoration could also be improved. Compat-
ible palettes yield convincing results, but dissimilar
palettes are integrated less well. Better color blend-
ing, especially in conjunction with oversegmentation,
is a direction for future work. The sign of the output
coefficient (positive or negative) is decided separately
from the magnitude and is a binary outcome. While
generally effective for static textures, the process has
a discontinuity as the two input values become simi-
lar in magnitude, which could be a problem for future
efforts on dynamic textures.
5 CONCLUSIONS AND FUTURE
WORK
We presented a method for synthesizing textures in-
termediate between two exemplars. The degree of
blending can vary spatially, yielding inhomogeneous
textures if desired. Using the SLIC segments we were
able increase heterogeneity, preserving small coher-
ent regions of each texture.
Our process is applicable only for textures lacking
well-defined structures. It also has difficulty preserv-
ing long continuous features. However, for stochastic
textures with small features, our output textures look
realistic and natural. In future work, we would like
to consider additional factors, including local con-
trast and directionality. We would like to extend the
method to work well for more structured textures. We
would also like to use time as a factor in order to cre-
ate dynamic textures.
In this work, we treated all levels of the Laplacian
pyramid the same way. However, it might make sense
to investigate different treatments for different levels.
Depending on the texture, one level or another may
contain more of the structural content, and account-
ing for this in the merging could produce still bet-
ter blends. The example of Doyle and Mould (Doyle
and Mould, 2018) is instructive, albeit not in a texture
blending context.
ACKNOWLEDGEMENTS
This work was supported by the Natural Sciences and
Engineering Research Council of Canada. We thank
members of the Graphics, Imaging, and Games Lab at
Carleton University for many useful discussions dur-
ing the development of this work. Finally, thanks to
the reviewers for helpful comments.
REFERENCES
Achanta, R., Shaji, A., Smith, K., Lucchi, A., Fua, P., and
S
¨
usstrunk, S. (2012). SLIC superpixels compared to
state-of-the-art superpixel methods. IEEE Transac-
tions on Pattern Analysis and Machine Intelligence,
34(11):2274–2282.
Ashikhmin, M. (2001). Synthesizing natural textures. In
Proceedings of the 2001 Symposium on Interactive 3D
Graphics, I3D ’01, page 217–226, New York, NY,
USA. Association for Computing Machinery.
Bonneel, N., Peyr
´
e, G., and Cuturi, M. (2016). Wasserstein
barycentric coordinates: Histogram regression using
optimal transport. ACM Trans. Graph., 35(4).
Cook, J. D. (2010). Soft maximum.
Doyle, L. and Mould, D. (2018). Augmenting photographs
with textures using the Laplacian pyramid. The Visual
Computer, pages 1–12.
Efros, A. A. and Freeman, W. T. (2001). Image quilting for
texture synthesis and transfer. Proceedings of SIG-
GRAPH 2001, pages 341–346.
Efros, A. A. and Leung, T. K. (1999). Texture synthesis by
non-parametric sampling. In ICCV, volume 2, pages
1033–1038.
Gatys, L., Ecker, A. S., and Bethge, M. (2015). Texture syn-
thesis using convolutional neural networks. In NIPS
28, pages 262–270. MIT Press.
Heeger, D. J. and Bergen, J. R. (1995). Pyramid-based tex-
ture analysis/synthesis. In SIGGRAPH, SIGGRAPH
’95, page 229–238, New York, NY, USA. ACM.
Heitz, E. and Neyret, F. (2018). High-performance by-
example noise using a histogram-preserving blending
operator. Proc. ACM Comput. Graph. Interact. Tech.,
1(2).
Lockerman, Y. D., Sauvage, B., All
`
egre, R., Dischler, J.-
M., Dorsey, J., and Rushmeier, H. (2016). Multi-
scale label-map extraction for texture synthesis. ACM
Trans. Graph., 35(4).
Paris, S., Hasinoff, S. W., and Kautz, J. (2011). Local
Laplacian filters: Edge-aware image processing with
a Laplacian pyramid. ACM Trans. Graph., 30(4).
Sendik, O. and Cohen-Or, D. (2017). Deep correlations for
texture synthesis. ACM Trans. Graph., 36(4).
Zhang, J., Zhou, K., Velho, L., Guo, B., and Shum, H.-
Y. (2003). Synthesis of progressively-variant tex-
tures on arbitrary surfaces. ACM Trans. Graph.,
22(3):295–302.
Zhou, Y., Shi, H., Lischinski, D., Gong, M., Kopf, J., and
Huang, H. (2017). Analysis and controlled synthesis
of inhomogeneous textures. Computer Graphics Fo-
rum (Proceedings of Eurographics), 36(2):199–212.
Zhou, Y., Zhu, Z., Bai, X., Lischinski, D., Cohen-Or, D.,
and Huang, H. (2018). Non-stationary texture syn-
thesis by adversarial expansion. ACM Trans. Graph.,
37(4).
GRAPP 2021 - 16th International Conference on Computer Graphics Theory and Applications
168