AN AUTOMATED IMAGE-ANALYSIS SYSTEM FOR MAJOR MACERAL GROUP-ANALYSIS IN COALS

An image analysis system has been developed capable of performing major maceral group analysis, based on morphological features as well as grey scale shading. The technique can be used to carry out the analysis without significant input from the operator. The main advantage of the system lies in its ability to distinguish liptinite from resin. Previous grey scale methods have been unable to do this, especially with lower rank coals. The method is described, together with the settings of the appropriate parameters. The repeatability of the method is tested and found to be comparable with manual point counting. Five coals from different parts of the world, with different petrologies, were analysed by point counting and by image analysis and there is good agreement between the results.