Artificial vision system for the online characterization of the particle size distribution of bulk materials on conveyor belts

Abstract In this work, a methodology is presented to develop an artificial vision system to characterize the particle size distribution of granular products in motion on a conveyor belt. This methodology exploits the wealth of information stored in videos of the bulk material and, in particular, the information on the particle size distribution is extracted by multivariate and multiresolution image texture analysis. The method is applied to the case of a granulated microcrystalline cellulose directly discharged from a hopper on a conveyor belt. The effectiveness of the presented methodology to estimate the particle size distributions is demonstrated in terms of estimation accuracy, velocity, reliability, and non-invasiveness of the system.