IMAGE TEXTURE ANALYSIS BASED PROCEDURE TO CHARACTERIZE AND RECOGNIZE COAL MACERALS

Recurrences and distributions of macerals play a decisive role in coal beneficiation, since macerals have markedly different chemical and ph ysical surface characteristics which affect the floatability and the e lectrical separability of particles coming from comminution. Procedure s have been developed for the automatic recognition of macerals on the basis of their micro-textural characteristics, evaluated by means of image-processing techniques. In such a way the textural characteristic s of various macerals were recognized and a pattern vector was identif ied for each type of maceral. The procedure takes into consideration t he data acquired by means of an optical microscope using reflected lig ht on polished specimens of coal ores.