Automatic flotation froth bubble size distribution estimation using mean shift and watershed transforms.

Abstract

Flotation is a process used in the mineral industry toconcentrate valuable minerals. The efficiency of the processis influenced by among other the bubble size distribution ofthe flotation froth. In this investigation, we propose a newmethod for automatic estimation of the bubble sizedistribution using the mean shift algorithm and watershedtransform. The mean shift algorithm is used to find pixelclusters of particular modes of the probability densityfunction of the image data. Pixel clusters of which themodes are above a given threshold are used as markers forthe watershed transform. Experimental results show thatthe proposed method was fast and robust for determiningthe size distribution of different classes of flotation froth.