Texture Synthesis Using Single-Scale / Multiple-Scale Markov Random Field Models

Abstract: 
This fast multigrid colour texture synthesis algorithm starts with spectral factorization of an input colour texture image using the Karhunen-Loeve expansion. Single orthogonal monospectral components are further decomposed into a multi-resolution grid and each resolution factors are independently modeled by their dedicated Markov random field model. Finally single synthesized monospectral single-resolution texture factorss are collapsed into the fine resolution images and using the inverse Karhunen-Loeve transformation we obtain the required colour texture.

 

The multiresolution algorithm.

Examples

Natural textures (upper row) and their synthetic counterparts.
 
The armchairs scene (left) and its virtual model covered with synthetic upholstery, wood, wallpaper and carpet colour textures.

 
    Distributed texture synthesis using a single-scale / multiple-scale MRF models (requires enabled Java applets).
  Animated version of the web distributed modification of the algorithm.
Reference: 
Haindl, M., and V. Havlíček, "Multiresolution colour texture synthesis", Proceedings of the 7th International Workshop on Robotics in Alpe-Adria-Danube Region, Bratislava, ASCO Art, pp. 297-302, June, 1998.