Texture synthesis is the process of algorithmically constructing a large digital image from a small digital sample image by taking advantage of its structural content. It is object of research to computer graphics and is used in many fields, amongst others digital image editing, 3D computer graphics and post-production of films.
Texture synthesis can be used to fill in holes in images (as in inpainting), create large non-repetitive background images and expand small pictures. See "SIGGRAPH 2007 course on Example-based Texture Synthesis" for more details.
Contents |
"Texture" is an ambiguous word and in the context of texture synthesis may have one of the following meanings:
Texture can be arranged along a spectrum going from stochastic to regular:
These extremes are connected by a smooth transition, as visualized in the figure below from "Near-regular Texture Analysis and Manipulation." Yanxi Liu, Wen-Chieh Lin, and James Hays. SIGGRAPH 2004
Texture synthesis algorithm are intended to create an output image that meets the following requirements:
Like most algorithms, texture synthesis should be efficient in computation time and in memory use.
The following methods and algorithms have been researched or developed for texture synthesis:
The simplest way to generate a large image from a sample image is to tile it. This means multiple copies of the sample are simply copied and pasted side by side. The result is rarely satisfactory. Except in rare cases, there will be the seams in between the tiles and the image will be highly repetitive.
Stochastic texture synthesis methods produce an image by randomly choosing colour values for each pixel, only influenced by basic parameters like minimum brightness, average colour or maximum contrast. These algorithms perform well with stochastic textures only, otherwise they produce completely unsatisfactory results as they ignore any kind of structure within the sample image.
Algorithms of that family use a fixed procedure to create an output image, i. e. they are limited to a single kind of structured texture. Thus, these algorithms can both only be applied to structured textures and only to textures with a very similar structure. For example, a single purpose algorithm could produce high quality texture images of stonewalls; yet, it is very unlikely that the algorithm will produce any viable output if given a sample image that shows pebbles.
This method, proposed by the Microsoft group for internet graphics, is a refined version of tiling and performs the following three steps:
The result is an acceptable texture image, which is not too repetitive and does not contain too many artifacts. Still, this method is unsatisfactory because the smoothing in step 3 makes the output image look blurred.
These methods, such as "Texture synthesis via a noncausal nonparametric multiscale Markov random field." Paget and Longstaff, IEEE Trans. on Image Processing, 1998, "Texture Synthesis by Non-parametric Sampling." Efros and Leung, ICCV, 1999, "Fast Texture Synthesis using Tree-structured Vector Quantization" Wei and Levoy SIGGRAPH 2000 and "Image Analogies" Hertzmann et al. SIGGRAPH 2001. are some of the simplest and most successful general texture synthesis algorithms. They typically synthesize a texture in scan-line order by finding and copying pixels with the most similar local neighborhood as the synthetic texture. These methods are very useful for image completion. They can be constrained, as in image analogies, to perform many interesting tasks. They are typically accelerated with some form of Approximate Nearest Neighbor method since the exhaustive search for the best pixel is somewhat slow. The synthesis can also be performed in multiresolution, such as "Texture synthesis via a noncausal nonparametric multiscale Markov random field." Paget and Longstaff, IEEE Trans. on Image Processing, 1998.
Patch-based texture synthesis creates a new texture by copying and stitching together textures at various offsets, similar to the use of the clone tool to manually synthesize a texture. "Image Quilting." Efros and Freeman. SIGGRAPH 2001 and "Graphcut Textures: Image and Video Synthesis Using Graph Cuts." Kwatra et al. SIGGRAPH 2003 are the best known patch-based texture synthesis algorithms. These algorithms tend to be more effective and faster than pixel-based texture synthesis methods.
In pattern-based modeling [1] a training image consisting of stationary textures are provided. The algorithm performs stochastic modeling, similar to the patch-based texture synthesis, to reproduce the same spatial behavior.
The method works by constructing a pattern database. It will then use multi-dimensional scaling, and kernel methods to cluster the patterns into similar group. During the simulation, it will find the most similar cluster to the pattern at hand, and then, randomly selects a pattern from that cluster to paste it on the output grid. It continues this process until all the cells have been visited.
Realistic textures can be generated by simulations of complex chemical reactions within fluids, namely Reaction-diffusion systems. It is believed that these systems show behaviors which are qualitatively equivalent to real processes (Morphogenesis) found in the nature, such as animal markings (shells, fishs, wild cats...).
Some texture synthesis implementations exist as plug-ins for the free image editor Gimp:
A pixel-based texture synthesis implementation:
Patch-based texture synthesis using Graphcut:
Several of the earliest and most referenced papers in this field include:
although there was also earlier work on the subject, such as
(The latter algorithm has some similarities to the Chaos Mosaic approach).
The non-parametric sampling approach of Efros-Leung is the first approach that can easily synthesis most types of texture, and it has inspired literally hundreds of follow-on papers in computer graphics. Since then, the field of texture synthesis has rapidly expanded with the introduction of 3D graphics accelerator cards for personal computers. It turns out, however, that Scott Draves first published the patch-based version of this technique along with GPL code in 1993 according to Efros.