Intensity-Based Image Segmentation (IBIS) for Flow Diagnostics in Suspensions

We present IBIS, Intensity-Based Image Segmentation, a novel method for robust particle image segmentation in optical flow diagnostics. The use of fluorescent tracers is the current state-of-the-art for flow disgnostics in suspension rheology, however, their implementation often incurs significant costs. While Refractive Index Matched (RIM) Particle Image Velocimetry (PIV) is a strong contender for flow visualization, the available sizes of RIM particles severely limits the parameter space. IBIS offers a cost-effective alternative, combining median and Gaussian filtering with adaptive binarization to effectively distinguish particles (Suspended solid particles and PIV tracers) based on their intensity and size characteristics. The algorithm proceeds by first subtracting background noise, then applying a series of filters to isolate distinct intensity features. Optimal performance is achieved by carefully adjusting camera aperture and frame exposure to ensure suspended particles are well-exposed while tracers do not saturate the sensor, maximizing the algorithm's efficiency in separating the two. This approach enables precise segmentation of both large and small particles, even in high-density flows. The solid particle concentration fields as well as the flow field around the suspension can be studied simultaneously, providing valuable insights into flow dynamics. IBIS offers a versatile and computationally efficient tool for analyzing complex multiphase flow phenomena.