Author: E.K. Algazinov, M.A. Dryuchenko, D.A. Minakov, A.A. Sirota, V.A. Shul'gin (Voronezh, Russia)
Measurement Techniques, New York, Vol. 57, Iss. 5, pp. 509-518 (Aug 2014)
DOI: https://doi.org/10.1007/s11018-014-0489-x
Abstract
A nonuniform flow of objects is analyzed based on processing of images synthesized from the results of line-by-line scanning of an illuminated flow in optical sorting systems. A hardware–software complex that functions in real time simultaneously with a sorting system in a combined mode of analysis of images of the flow based on proposed preliminary processing and recognition algorithms is described. Experimental results of these algorithms for the case of an analysis of a flow of components of grain mixtures are presented.
Keywords
Optical sorting, recognition of objects, image processing, neural networks
Full text