Quantitative characterisation of iron ore pellets with optical microscopy and machine learning

PelletisationRecycling and metallurgy


Ductus Preeye AB



Martin Simonsson



Purpose and goal
The goal was to evaluate the potential of a system of quantitative characterization of iron ore pellets based on automated microscopy, image analysis and machine learning. The system, after training with expert annotated data, could reliably identify a number of relevant phases like hematite, magnetite and metallic iron. The results matched well the manual assessment and contributed more information than previous systems. The potential for the new approach is therefore considered to be very large, and could contribute to job saving and increased knowledge.

Structure and implementation
Annotation of collected data was done in collaboration with experts, which ensured the quality of the training and test data. Some phases proved difficult to identify even for optical microscopy experts. Additional technologies such as SEM may be needed for this. Collected data can also be used to train and test new machine learning methods in the future, such as convolutional neural networks (CNN).

A relevant dataset with annotated microscopic images with iron ore pellets was created. This was then used to train and evaluate a number of classifiers. Based on this, relevant microstructures could be quantified as well as the amount and distribution of different phases, minerals and additives. In addition, porosity and size distribution on particles could be determined. With the help of collected and annotated data, the concept could be validated in a laboratory environment. Code optimisation and user interface development were assessed as prerequisites for a commercial system.