Texture analysis of brodatz images using statistical methods

Abstract EN

Textures are one of the important features in computer vision for many applications.

Most of attention has been focused on the texture features.

An important approach to region description is to quantify its texture content.

Although no formal definition of texture exists, intuitively this descriptor provides measures of properties such as smoothing and regularity.

The principal approaches used in image processing to describe the texture of an image region are statistical, structural, and spectral.

In this paper the features were constructed using different statistical methods.

These are auto-correlation, edge frequency, primitive-length and law’s method ; all these methods were used for texture analysis of Brodatz images.

The result showed that the law’s autocorrelation method yields the best result.