Sistema de navegação autônoma para o robô agrícola Soybot

Abstract

The autonomy of agricultural machines has been an important research field for Precision Agriculture. With the use of global positioning systems (GPS), the farmer is able to maximize the yield of his land through precise sowing. For other machines to perform actions after planting, the farmer often does not have data on the exact location of each row of the plantation; therefore, just being guided by the global position is not enough. The present work proposes an approach consisting of two steps for the control of the Soybot agricultural robot: a computer vision technique for the robot to travel through the interior of a grown plantation, without damaging it; and a path planning technique so that, at the end of the plantation row, it enters in the next row.