A new approach for froth image segmentation using fuzzy logic

Abstract

The segmentation of flotation froth images is extremely challenging mainly due to the dynamic and continuous nature of the process as well as large number of closely packed bubbles of different sizes with barely detectable boundaries. All these characteristics of the froth images have complicated the development of a reliable and comprehensive segmentation algorithm. In the present work, an adaptive fuzzy model based on multi-scale and multi-directional simulation of the human visual system is developed to delineate the bubbles in the flotation froth images. The implemented model is successfully validated using some laboratory and industrial scale froth images.