Diagnosis of concentrate grade and mass flowrate in tin flotation from colour and surface texture analysis

It has long been accepted that the visual appearance of a froth is generally a good qualitative indicator as to the performance of the flotation process. This paper describes work which has been undertaken to investigate the possibility of using this premise to develop an on-line control system for the flotation process using image analysis. erging technologies of image analysis and neural networks have been applied to images of flotation froths captured at an industrial tin flotation plant. The images were processed to produce a colour profile and textural descriptors of the froth surface. This data was then correlated against performance characteristics, such as concentrate grade, water and solids flowrates, using both conventional statistical and neural network modelling. shown that the colour of the froth can be used to give a prediction of the grade of the concentrate produced. Different procedures of describing the surface texture of the froth were studied, including a novel method using fractal analysis. It is shown that the texture of the froth is a good indicator to the performance of the flotation cells. The possibilities of using this in a robust control system are discussed.