this paper, an indirect method for determining the Sauter mean diameter of a bubble population is presented. The technique relies on the estimation of the percent area occupied by bubbles in a 2D image as well as on the superficial gas rate. The percent area was defined as the quotient between the black pixels (bubble representation) divided by the total pixels in a binary image. A linear model to describe the Sauter diameter as a function of the percent area and the superficial gas rate was proposed. The regressions showed a fairly good predictive capacity in laboratory and industrial flotation machines under bubbly regime, which makes the proposed methodology a promising tool for monitoring and controlling gas dispersion in real time. (C) 2015 Elsevier Ltd. All rights reserved.