Przemysl Chemiczny, Vol.82, No.8-9, 1275-1278, 2003
Automated classification of industrial fluids based on microscopic images
A fully automated procedure was developed to take microscopic images of impurity particles collected from industrial oils. New image interpretation algorithms were developed to determine the type, the quantity and the size of solid impurities (metal, resin-like, quartz, synthetic and natural fibers). The algorithm allowed to assign to the oil the cleanliness class in compliance with ISO 4406 and NAS 1638.