Identification of Agricultural Plant Row Using the Clustering Algorithm in the Navigation of Mobile Robot

Abstract

This paper describes navigation method usable in agriculture plant row for mobile robot. The proposed method with clustering algorithm uses data obtained by Hokuyo laser sensor. Using the created solution it is possible to determine plant in a row when mobile robot using the local navigation principle. Crops in a row are characteristic waypoints that create the vertices of the polygon. From this polygon the center of gravity is calculated, which determines the center position of the plant row. From this polygon the center of gravity is calculated, which determines the center position of the plant row. This position is used for mobile robot control algorithm. The processed data by scanners provide information not only on waypoints but also on any obstacles in the agricultural plant row.