Machine Vision System for Automated Detection of Stained Pistachio Nuts

A machine vision system was developed to separate stained pistachio nuts, which comprise about 5% of the California crop, from unstained nuts. Stained nuts have lower consumer acceptance and higher incidences of aflatoxin contamination. The machine vision system may be used as an automated quality control device to detect stained nuts. The system was tested on three different pistachio process streams: the bichromatic colour sorter reject stream (floaters and sinkers) and the manually sorted small nut shelling stock stream (floaters). The system had a minimum overall error rates of 13% for the sinker bichromatic reject stream, 14% for the floater bichromatic sorter reject stream and 15% for the floater small shelling stock stream.