Verfahren und Anordnung zur optischen Beurteilung von Erntegut in einer Erntemaschine

摘要

The method involves dividing an image by a binary gumption mask into single objects and a background, where the single objects are further processed. An application of the binary gumption mask is performed with sharper discrimination parameters such that the application of the binary gumption mask is split into individual objects and/or properties of the individual objects are identified and merged based on thresholds of the single objects. The properties e.g. color, outline and/or texture features, of the single objects are classified. An independent claim is also included for an arrangement for performing optical assessment of crop.