Developing a small autonomous robot for crop inspection.

Abstract

This paper presents the development and evaluation of an autonomous field inspection vehicle that navigates visually guided by the crops rows with the aim of crop inspection and weed mapping. Therefore, an image processing method capable of extracting in real time the crop rows from images taken with a quality web camera situated in the front of the robot has been developed. The aim of the image processing is to extract the vehicle’s relative position to the crop rows, i.e. the motion direction angle (the crop rows are parallel to each other) and displacement (offset) between the vehicle centre and the middle point between two crop rows. Determined values of the vehicle’s offset and angle are supplied to a fuzzy controller that calculates the correction values for the steering, in order to generate in the vehicle a crop row follow behaviour. Results show the correct performance of the autonomous robot developed.