Traditional quality control procedures for aggregates production within
the quarry industry rely on off-line laboratory analysis, resulting in poor
performance and inconsistent product quality. A high speed on-line
aggregate classification system is described here. A product sampling
system directs aggregate through an inspection chamber where a novel
imaging system digitises the particle geometry. Data processing
algorithms extract dimensional data and image features, while a neural
network classifier assigns a qualitative shape descriptor to each rock
particle. This system provides each sampled batch with a complete
analysis of size and shape distributions, complementing a new generation
of cone crusher technology to maximise product quality and plant
efficiency.