Pellet Sphericity and Size Analysis by Image Processing Technique

Abstract

Preparation of iron ore fines as feed to blast furnace involves production of spheres of typically 9-16mm diameter through pelletization. The process combines agglomeration and thermal treatment to convert the raw ore into pellets appropriate for use in a blast furnace. The use of pellets of varying shape and size has attracted much interest. The permeability of blast furnace is dependant on the size and shape of the pellets. The productivity of pelletization is based on generation of pellets of size range of 9-16mm. The pelletization process is governed by the physical characteristic of pellets and therefore it is necessary to investigate accurately the shape and size of the pellets in pelletization process. It is desirable to produce nearly spherical pellets with a desired diameter range. In the process of pelletization the pellets may be produced of different sizes. This may be accepted if it is within certain tolerance limit. Therefore, it is essential to measure the size and provide a quantitative value for the shape of the pellets. The size can be measured by many methods such as using a scale or caliper and solid displacement method etc. However, it is very difficult to measure the size and shape of green pellets without physically handling them. Since they are wet and they change their size and shape during measurement. Some researchers have worked using imaging techniques for pellet size analysis [3-4]. This motivated to develop such a system. An attempt is made here to measure the shape and size of the pellets using 2D-digital image processing techniques. Online image processing method is also presented for calculating the sphericity and the size of a single/multiple pellet by software and its toolboxes. The technique can suitably derive the size distribution of a number of pellets. The technique is validated by conducting a number of experiments