Coal structure characterization and analysis by digital texture processing

Abstract

Coal is a non-homogeneous substance; it consists of different constituents called macerals, each one characterized by its own set of properties. Macerals differ from mineral constituents of rocks because they do not have a well defined chemical composition. The purpose of this paper is to describe the use of digital processing techniques based on the analyses of sample images to characterize coals taking into account the reflectance-textural characteristics of macerals. These characteristics were recognized, analysed and modelled. A pattern vector characteristic of each maceral was thus identified. The procedure is based on the processing of data (digital images) acquired by means of reflected light optical microscopy. Polished specimens of coal ores are considered for the study. In such a way it is pointed out a suitable procedure for a quantitative characterization of the different macerals constituting the coal.