Solid-texture synthesis: a survey

It's common belief that textures can simply and efficiently model 3D objects by separating appearance properties from their geometric description. Computer graphics has profusely used textures to model objects' external structure, through either photographs or procedural models.1 Whereas traditional 2D textures usually encode information about an object's external surface, researchers have proposed extensions for providing volumetric information, allowing encoding of objects' internal appearance. That is, these extensions provide appearance properties for each point in a predefined volumetric domain D ⊂ R3. Such textures are usually called solid textures. This survey illustrates the different algorithms for synthesizing and representing these textures