Publication: Autonomous weeding robot based on image processing using labview
Loading...
Date
2020-08-01
Authors
Vicky Kong, Wei Hau
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Weeds are unwanted wild plants that grow in the midst of crops to compete with them for nutrients, sunlight, water etc. In precision agriculture, the practice of spraying herbicide is often implemented to control the growth of weeds in nature. Apart from crops field, weeds also exist in house lawns. It is undeniable that the use of herbicide can cause soil pollution and endanger the neighbouring grasses. Besides, it is also time-consuming to remove all the weeds by hand. In order to tackle these challenges, an autonomous weeding
mobile robot was designed to pull the weeds out of lawn. The weeding robot consists of a vision system, gripper mechanism and mobile platform. The vision system includes a
webcam mounted on the mobile platform to capture the real-time image in lawn. The image processing algorithm for weeds identification in lawn was designed using LabVIEW and
the four-wheeled robot was controlled using Arduino Mega. The gripping mechanism on the other hand was designed using SolidWorks and printed using 3D printer. The
performance analysis of the weeding robot was carried out on the artificial lawn by allowing the robot to move in a straight path and remove the weeds along the way. The result showed an average success rate of 90% for weeds identification and picking after 5 attempts.