Open AccessIdentification of transition processes parameters in flotation machine by computer vision
Author(s) -
A.V. Zatonskiy
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/860/1/012021
Subject(s) - frame (networking) , process (computing) , computer science , luminosity , identification (biology) , contrast (vision) , bubble , artificial intelligence , computer vision , control (management) , process engineering , algorithm , control theory (sociology) , simulation , engineering , physics , astrophysics , telecommunications , botany , galaxy , parallel computing , biology , operating system
The efficiency of potassium fertilizers production needs to improve industrial control systems because the fertilizers are very important for agriculture and economy all over the world. Now a control of flotation quality is made by a visual operator’s observation, and a human factor could influence on the results. This paper is about of researching of foam recognize algorithms based on determination of flares from bubble surfaces. Statistics of flares are used for determination of parameters of transition processes in the flotation machine. A method based on bubble flares detection instead of bubble borders recognition. The method allows recognizing bubbles even on a surface of specific potassium foam, non-contrast, solid and color. This research solves the tasks of an optimal binarization threshold approximation, foam driver imagination exclusion, transition process detection. The optimal threshold is developed from medium frame luminosity quite linear. The foam driver can be excluded from the frame by specific character of luminosity trend. The transition process leads to change of bubbles count easy detected by software. A time of processing is about 15-30 ms per frame. Thus, the method can be used as in alarm systems, so in control systems on decision support systems.