Automatic classification of olives for oil production using computer vision

One of the most important parameters in olive oil production is the correct reception and classification of olive fruits in batches before starting the oil extraction process. This work proposes an automatic inspection system based on computer vision to classify automatically different lots of olives for oil production when the milling process starts. The classification is based on the differentiation, on line, between ground and tree olives.

For this purpose, the samples of olives have been obtained by picking berries directly from the tree or from the ground. The feature vector of the samples has been obtained on the basis of the olive image histograms. Moreover, different image processing algorithms have been employed and lineal classification techniques such as Fisher Discriminant Analysis. The system has reached good classification results distinguishing between soil and tree olives batches with success ratios of 100%.