Development and Validation of an Online Analyzer for Particle Size Distribution in Conveyor Belts

Online measurement of particle size distribution in the crushing process is critical to reduce obstruction and energy consumption. Nevertheless, commercial systems determine do not accurately identify large particles (20–250 mm), leading obstruction, increasing consumption, reducing equipment availability. To solve this problem, an online sensor prototype was designed, implemented, validated a copper ore plant. The based on 2D images specific detection algorithms. system consists camera (1024 p) mounted conveyor belt image processing software, which improves edges. algorithms geometry each particle, from sequence digital photographs. For development noise reduction were evaluated selected, routine designed incorporate morphological mathematics (erosion, dilation, opening, lock) segmentation (Roberts, Prewitt, Sobel, Laplacian–Gaussian, Canny, watershed, geodesic transform). software implemented (in MatLab Image Processing Toolbox) 3D equivalent diameter (using major minor axes, assuming oblate spheroid). adjusted Rosin Rammler function axis. test capabilities, laboratory used, where results show precision 5% Rambler model fitting. validate algorithms, pilot mining company Chile. accuracy 60% 67% depending stage. In conclusion, it shown that allow sizes, provides useful information for screening maintenance control crushers’ open setting, risk operational