COAL PETROGRAPHY - A PATTERN-RECOGNITION APPROACH

A pattern recognition based technique has been used to classify the di fferent constituents (macerals) of coal. The method has two modules, a n off-line training module and an on-line classification module. Three -band color (R-G-B) images of coal samples are used in contrast to mon ochrome (B and W) images used by earlier researchers. Each maceral cla ss is determined based on the gray values (or the reflectance properti es) in the three bands. Points are assigned to one of vitrinite, exini te, inertinite or background class using a minimum distance classifier . This means that the distance of a point from its assigned cluster ce ntre is minimum with respect to the distance of the point from other c luster centres. This method also minimizes the sum of squared error of mis-classification. The method has a quantitative basis to standardiz e petrological analysis of coal samples in relation to reflectance pro perties of different macerals under a given lighting condition. The sp eed of the system is considerably faster than the current technique of analysis and it ensures repeatability of results.