3D Markov random field binary texture model: Preliminary results

Texture analysis and synthesis is an important field in computer vision. Since the late sixties, numerous techniques have been developed for the synthesis and analysis of texture in 2D images. However, only a few models have been presented to synthesize 3D texture images and in most cases 2D texture mapping is used to emulate this process. In addition, most of the techniques used for the analysis of texture in 3D medical images, such as CT and MRI, are based on 2D models applied to each slice followed by reconstruction of the volume. This approach does not use all available information contained in the data. A more robust solution is given by solid texture modelling. The paper describes a novel approach to solid texture modelling based on Markov random fields