Abstract

Research and development into the application of machine vision in froth flotation systems has continued since its introduction in the late 1980s. Machine vision is able to accurately and rapidly extract froth characteristics, both physical (e.g. bubble size) and dynamic (froth velocity) in nature, from digital images and present these results to operators and/or use the results as inputs to process control systems. Currently, machine vision has been implemented on several industrial sites worldwide and the technology continues to benefit from advances in computer technology. Effort continues to be directed into linking concentrate grade with measurable attributes of the froth phase, although this is proving difficult. As a result other extracted variables, such as froth velocity, have to be used to infer process performance. However, despite more than 20 years of development, a long-term, fully automated control system using machine vision is yet to materialise. In this review, the various methods of data extraction from images are investigated and the associated challenges facing each method discussed. This is followed by a look at how machine vision has been implemented into process control structures and a review of some of the commercial froth imaging systems currently available. Lastly, the review assesses future trends and draws several conclusions on the current status of machine vision technology.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call