The automation of intuitive visual expertise: classification of beef carcasses using neural networks

Throughout Europe beef carcasses are classified by visual inspection according to the so-called EUROP scheme regulated by the European Union. The EUROP scheme defines fatness and conformation (shape) on scales from 1 to 15. This is the basis for payment to the farmer. The classification is carried out by the meat plant´s classifier-a person trained to grade according to the EUROP scheme. The classifiers are controlled by government inspectors, which in turn are controlled by a board of European inspectors. The paper describes a computer vision system using neural networks which is able to outperform and replace the classifier