Leading point based interrow robot guidance in corn fields

Abstract

Various approaches of vision based techniques has been implemented for mobile robot guidance in corn field. The aim of robot control is to keep the robot stay on interrow center line. This paper presents a new method of vision-based mobile robot guidance between rows of medium size corn plants when the field of view is not parallel to ground level and view of corn rows is incomplete. Instead of plant oriented technique, the proposed method is focusing on soil called Leading point based guidance. Method of the leading point search in image and how the leading point data is applied on Fuzzy Logic Design for robot PID heading control are explained. Different interrow images of corn field are tested. Results of image processing show success of the proposed leading point search method and simulation demonstrates differential drive robot controlled by the leading point data.