Flotation froth monitoring using multiresolutional multivariate image analysis

MR-MIA is applied to RGB color images of the froth in a zinc recovery section of Agnico-Eagle's Laronde plant in Quebec to extract textural and color information related to the bubble size distribution, and the presence and amount of clear windows or black holes in the froth. A principal component analysis (PCA) model built using this information on the froth structure then is used to provide a compact representation of the health of the froth. Process monitoring plots are then presented to track the dynamic variations in the froth health over a range of operating conditions.