Characterization of coal particles using rule base on-line image analysis. Final technical report, September 1, 1993--August 31, 1994

Abstract

The degree of pyrite liberation is a key to determine whether a batch of ground coal obtained through a grinding process is acceptable. In this regard, a synthesis so-called on-line image analysis system to be used as a part of a coal cleaning process has been developed. This system is designed to replace the traditional method in the assessment of pyrite liberation of ultrafine coal. The online image analysis system consists of two parts: an automated image analysis computer program and an automated sample handling system which is an assembly of an optical microscope, a video camera, a X-Y table, an autofocus unit, and a personal computer. The automated image analysis computer program performs the analysis of coal particles without any intervention of a human operator. The image analysis computer program is a rule base decision-making algorithm and is the creation of an extensive study of image characteristics of coal particles such as shade of gray, particle size etc. In general, the image analysis computer program developed in this project can identify different particles in a coal image as good as a human operator does. On the other hand, the sample handling mechanisms are controlled through the interaction between the step motors and an feedback unit attached to a personal computer. Naturally, a computer program is needed in order to send commends to move the X-Y table. For example, the control of the autofocus mechanism was achieved by comparing the development of particle edge in the images of different focal planes. It was found that an image is brought to focus when most of the edges is detected. In summary, the findings of this project have led to believe that the on-line image analysis can be used as a part of the coal cleaning process.