Methods and Algorithms for Segmentation of the Image in Control Flotation Process

The aim of the study is to develop the methodology for building a granulometric histogram of the foam product of the flotation process. One of the stages of the technique is the construction of an algorithm for image segmentation based on the glares of bubbles and the determination of their area. A feature of this algorithms is the use of a binarized image with a color threshold the histogram is insensitive to, and the possibility of estimating the area of segments. For the developed algorithms the program implementation in C++ has been performed. Results of testing algorithms for binary images with different color thresholds are given. The obtained results showed that on the basis of the analysis of the glare sizes of the bubbles of the foam product, a granulometric evaluation of the foam product of a flotation machine can be performed.