An Efficient Topological Characterization of Gray-Levels Textures, Using a Multiresolution Representation
✍ Scribed by Arie Pikaz; Amir Averbuch
- Publisher
- Elsevier Science
- Year
- 1997
- Tongue
- English
- Weight
- 850 KB
- Volume
- 59
- Category
- Article
- ISSN
- 1077-3169
No coin nor oath required. For personal study only.
✦ Synopsis
function or by a vector of features. In this paper we present a new method for texture characterization, which is based For a given textured image we define a sequence of graphs ͕N s (t)͖ sʦI (called MRCG), where N s (t) is the number of 4-con-on topological properties at different gray-levels and differnected components with size Նs, for the images thresholded ent resolutions. We use a data-structure, that was first with t. The sequence of graphs are computed in almost linear represented in [19,20], named MRCG (multiresolution time complexity, where the input size is the number of pixels clusters graphs). The MRCG data structure is composed of the image, using only integer arithmetic. The MRCG is a of a sequence of graphs, denoted by ͕N s (t)͖ sʦI . For a given multiresolution representation, which is very useful for texture s and t, N s (t) is defined as the number of connected objects characterization. Its properties as a texture characterizer are which are composed of at least s pixels, if the image was analyzed and demonstrated.