Robotic intra-row weed hoeing in maize and sugar beet

Abstract

A prototype of robotic intra-row weed hoeing in maize and sugar beet is presented in this study. Weeds in the crop rows were identified using a bi-spectral image analysis system and shape analysis. Positions of weeds in the images were recorded. Selective weed control in the row was performed with a modified finger weeder driven by electrical motors. Speed of the finger weeder was increased at positions where only weeds were classified. The system was triggered by an encoder and controlled by a micro-controller.