Froth image segmentation algorithms and their validation

Abstract

In mineral processing, froth characteristics are important for flotation results. To monitor the variation of froth characteristics, an optical system has been developed. In the flotation system, one important procedure for image processing and analysis is image segmentation. This paper presents four newly developed segmentation algorithms for froth images. Results of image segmentation were compared to manual segmentation results, and comparisons show that the newly developed algorithms work well for the tested froth images.