Texture synthesis

From Wikipedia, the free encyclopedia

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.

In few words, texture synthesis is used to fill in holes in images, create large non-repetitive background images and expand small pictures.

1 Textures
2 Aim
3 Methods
4 Literature
5 External links
6 References

[edit] Textures

Maple burl, an example of a texture.

"Texture" is an ambiguous word and in the context of texture synthesis may have one of the following meanings:

In common speech, "texture" is a synonym for "surface structure". Usually it comprises the five properties attributed to texture by psychology of perception: coarseness, contrast, directionality, line-likeness and roughness ^[1].
In 3D computer graphics, a texture is a digital image that is applied to a three-dimensional model via texture mapping to give objects a more realistic look. Often, such an image is a photograph of a "real" texture, such as wood grain.
In image processing, every digital image composed of repeated elements is called a "texture." For example, see the images below.

Texture can be arranged within a spectrum from Stochastic to Regular:

Stochastic textures. Texture images of stochastic textures look like noise: colour dots that are randomly scattered over the image, barely specified by the attributes minimum and maximum brightness and average colour. Many textures look like stochastic textures when viewed from a distance. An example of a stochastic texture is roughcast.
Structured textures. These textures look like somewhat regular patterns. An example of a structured texture is a stonewall or a floor tiled with paving stones.

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

[edit] Aim

The aim of a texture synthesis algorithm is to create an output image that meets the following requirements:

The output shall have the size determined beforehand by the user.
The output shall be as similar as possible to the sample.
The output shall not have visible artifacts such as seams, blocks and misfitting edges.
The output shall be non-repetitive, i. e. there shall be no identical structures in the output image.

Moreover, it shall meet the usual requirements of algorithms: efficiency in terms of speed and economy in terms of memory.

[edit] Methods

The following methods and algorithms have been researched or developed for texture synthesis:

[edit] Tiling

The plainest way to create a larger image from a sample image is by simply tiling it. That means the output image is tiled together from duplicates of the sample image, in a simple copy-and-paste manner. The result is usually unsatisfactory: unless the sample complies with special requirements, the seams between the tiles will be plainly visible, and even then still the output image will be highly repetitive.

[edit] Stochastic texture synthesis

An algorithm of that family produces an image by randomly chosing 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.

[edit] Single purpose structured texture synthesis

Algorithms of that family use a fix 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.

[edit] Chaos mosaic

This method, proposed by the Microsoft group for internet graphics, is a refined version of tiling and performs the following three steps:

The output image is filled completely by tiling. The result is a repetitive image with visible seams.
Randomly selected parts of random size of the sample are copied and pasted randomly onto the output image. The result is a rather non-repetitive image with visible seams.
The output image is filtered to smooth edges.

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.

[edit] Pixel-Based Texture Synthesis

These methods, such as "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.

[edit] Patch-Based Texture Synthesis

Patch-based texture synthesis creates a new texture by copying and stitching together textures at various offsets. "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.

An example of Efros and Freeman's Image Quilting.

[edit] Literature

Several of the earliest and most referenced papers in this field include:

Popat in 1993 - "Novel cluster-based probability model for texture synthesis, classification, and compression".

Heeger-Bergen in 1995 - "Pyramid based texture analysis/synthesis".

Efros-Leung in 1999 - "Texture Synthesis by Non-parameteric Sampling".

Wei-Levoy in 2000 - "Fast Texture Synthesis using Tree-structured Vector Quantization"

although there was also earlier work on the subject, such as

Gagalowicz and Song De Ma in 1986 , "Model driven synthesis of natural textures for 3-D scenes",
Lewis in 1984, "Texture synthesis for digital painting".

(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.

A similar technique for audio rather than images is known as Granular synthesis.