Kartoffel Ertrags Kartierung durch Bildanalyse Potatoe Yield Mapping by Machine Vision

摘要

From the work presented in this paper can be concluded that there are good perspectives for a machine vision based on-line potato yield mapping system using 2 D potato information. For this moment the average error for potatoes free of tare is estimated between 4 and 5% and can be reduced when further system improvements are implemented. Software modification will enable unlimited cluster separation. In field experiments the system has to be validated. Additional research is necessary to handle soil covered tubers.