It is well recognized that flotation is a multivariate
process influenced by many factors such as reagent doses
of chemicals, aeration rate, and impeller speed. The
modeling and the control of flotation processes are
challenging due to the inherently chaotic nature of the
underlying microscopic phenomena. At present, the
control of flotation process depends heavily on various
experience of human operators by viewing the visual
appearance of the froth. Based on the fact that vision
information of froth layer surface is of great importance
to the flotation process monitoring and control[1], it is a
breakthrough to choose digital image processing as a
novel tool to gain a better understanding of the industrial
flotation process.