Flotation froth images velocity feature extraction and analysis based on Fourier-Mellin transform and gray-template matching

The variable of froth velocity is an effective flotation optimal control parameter in the machine vision based flotation monitoring and control, which closely related to the flotation performance indices. However, the bubble velocity extracted based on digital image processing is often difficult to accurately extract, due to the inevitable rotation, scaling and the collapse of the bubbles deformation and so on. The problem that hard to extract the velocity characteristic of the deformed bubbles was solved by applying the gray template matching fused Fourier-Mellin transform based registration, hence the velocity distribution of the bubble images was extracted accurately in terms of the feature of the bubbles. Then, the relationship between the bubbles velocity turbulence degree and the flotation performance indices was discussed. The method to extract the bubble velocity has been applied on the flotation field of industrial scale for more than two years, which provides guidance to the operator through extracting the velocity field and observing the chaos level of the bubble motion velocity of the bubble images captured in real time automatically.