|
| |
Texture
Chapter 14.1 Statistical texture description
14.2 Syntactic texture description

Overview
 |
There is no formal definition of texture. However it is
usually interpreted to measure qualities such as smoothness, coarseness,
regularity, and direction |
 |
Texture can be characterized by different metrics such as:
 |
statistical metrics: characterize smooth, coarse,
grainy, etc. using traditional statistical values such as mean and
variance. |
 |
structural metrics: patterns based on image
primitives such as lines or other repeating structures in the image |
 |
spectral metrics: based on properties of the
Fourier Transform such as symmetry, directionality. |
|
 |
Applications of Texture Processing
 |
Segmentation |
 |
Create textures for image generation, compression,
steganography (image watermarking) |
|


Methods based on spatial frequencies



Co-occurrence matrices












Edge frequency




Primitive length (run length)








Laws' texture energy measures



Fractal texture description










Other statistical methods of texture
description


Shape chain grammars




Graph grammars

Primitive grouping in hierarchical
textures






Page last updated on 04/06/01
|