A novel image analysis solution based on multiresolutional multivariate image analysis (MR-MIA) is proposed for the monitoring and control of flotation processes. The approach is quite different from the contemporary image analysis approaches in the sense that it can handle spatial (i.e., morphological) and color information of froth images efficiently, and is inherently robust to image quality and lighting conditions. 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. (C) 2004 Published by Elsevier Ltd.