producing a multiscale texture synthesis

"Texture synthesis via a noncausal nonparametric multiscale Markov random field ..

Grinspun, “Multiscale texture synthesis ..

Theresulting adaptive representations are discrete, orthogonal and allow amultiscale description of the geometric content of an image.

This leads to the construction of orthogonal bandelet bases, for whichthe grouping process is locally defined using a best orientation. These orthogonal bases improve over state of the art schemes for imagesand surfaces compression and for the inversion of the tomographyoperator.

In order to understand and model the complex geometry of turbulenttextures, we design an association field that is able to capture longrange interactions. This allows a statistical modelling of thegeometry of natural textures. We apply this construction togeometric texture synthesis.

X. Qin and Y.H. Yang,

Optimization-Based Multi-scale Texture Synthesis - …

In summary, we feel that multiscale texture synthesis is simple in concept but powerful in application; it directly provides novel and useful applications, while introducing a host of new avenues f...

Image analogies are based on asimple multiscale autoregression, inspired primarily by recent resultsin texture synthesis.

Other interesting ideas include the synthesis of multiscale textures directly on large meshes, or in application to solid texture synthesis [=-=Kopf et al.

In this paper, we develop a multiscale texture synthesis algorithm