Texture metric of halftone images

Abstract

Texture plays an important role in image features, in both display and perception of halftone images. We propose a quantitative measure for the occurrence of textures in binary images. This metric considers the fact that textures may be disturbed by missing or spurious pixels and that texels may overlap. Moreover, the texture may be black on white or white on black, which is also considered in the texture metric. We analyze various binary images with different textural characteristics with respect to different textures, to demonstrate the properties of the texture metric.