IMPROVEMENT OF COMPUTER RECOGNITION OF FOAM PARAMETERS IN POTASH FLOTATION MACHINES BY ANTI-GLARES FROM BUBBLES

Abstract

The article is about a problem of improving the quality of foam layer parameters recognition in potash flotation machines. The urgency of the problem has recently been emphasized by the need to strengthen Russia's food security, for which, in particular, potash fertilizers are required. The glare method of recognizing the parameters of the foam layer in potash flotation machines makes it possible to eliminate the human factor from control, improve product quality and reduce waste. However, it has a number of disadvantages – a large spread of results, sensitivity to the organization of lighting, etc. The aim: to improve the quality of foam layer parameters recognition by taking into account not only glare, but also anti-glare from bubbles. Materials and methods. Anti-glare is recognized by two-level adaptive binarization of the image. Dozens of video sequences were processed from slurry, silvin and pneumatic ejector flotation machines on two potash enterprises. The possibility of linear approximation of not only the upper one, but also the lower level of binarization is investigated. Binarization profiles for additional recognition of glare and anti-glare are constructed. Results. The possibility of using anti-glare bubbles to improve their recognition by an average of 3.6% is shown. The greatest increase in quality is achieved on pneumatic ejector flotation machines. The estimation of the average deviation suggests that the accuracy of alarm deviations due to noise reduction will improve by 1.4–4.8%. The anti-glare profiles character shows that a small shift in the adaptive level of binarization can further improve recognition in the range of 0.6–1.5%. The characteristic features of the method are revealed, for example, its superiority over the original glare method in the sense of a lower sensitivity to the error of calculating the binarization level. The double increase in recognition time is a disadvantage of the method, but it does not prevent one’s using, including using of multi-frame noise filtering. Conclusion. The conducted research suggests that taking into account the anti-glare bubbles when recognizing the parameters of potassium foam flotation machines can improve the process performance and can be applied in practice, preferably for pneumatic ejector machines.