|
| |
|
|
Dynamic Textures Modelling with Temporal Mixing Coefficients Approximation
Havlíček, Michal
Appearance of many real world materials is not static but changes in time. In case of spatially and temporally homogeneous changes the material can be represented by means of dynamic texture. Dynamic texture modelling is a challenging problem. In this article we present possible solution based on eigen analysis of input data and subsequent processing and modelling of temporal interpolation eigen coefficients using a combination of piecewise linear approximation and normal distribution sampling. The proposed method shows good performance, enables compress significantly the original data and extremely fast synthesis of arbitrarily long extension of the original texture.
|
|
|
MSAR BTF Model
Havlíček, Michal
The Bidirectional Texture Function (BTF) is the recent most advanced representation of material surface visual properties. BTF specifies the changes of visual appearance due to varying illumination and viewing conditions. Such a function might be represented by thousands of images of surface taken in given illumination and viewing conditions per sample of the material. Resulting BTF size, hundreds of gigabytes, excludes its direct rendering in graphical applications, accordingly some compression of these data is obviously necessary. This paper presents a novel probabilistic model based algorithm for realistic multispectral BTF texture modelling. This complex but efficient method combines several multispectral band limited spatial factors and corresponding range map to produce the required BTF texture. Proposed scheme enables very high BTF texture compression ratio and in addition may be used to reconstruct BTF space i.e. non-measured parts of the BTF space.
|
|
|
Near-periodic texture synthesis
Hatka, Martin
This paper describes two methods for seamless enlargement of difficult colour textures containing both regular periodic and stochastic components. Such textures cannot be modelled using neither simple tiling nor using purely stochastic models. The first novel method automatically recognizes and separates periodic and random texture components. Each of this components is subsequently modelled using the coresponding optimal method. Both independently enlarged texture components are combined in the resulting synthetic near regular texture. The second method detects two main direction of periodicity in regular periodic component and generates several double-toroidal tiles of the same general shape, which can seamlesly enlarge given near-periodic texture without visible regularity.
|
|
| |
|
| |
|
| |
guest ::
