An intelligent spraying robot based on plant bulk volume

Abstract

Spraying in farmlands and greenhouses is not only labor and time consuming, but also it has adverse effects on human health. The best solution is to use mechanized devices and robots to automate the spraying operation. In this study, a robot with four degrees of freedom performs spraying through calculating the plant bulk volume. The robot travels along the plant rows and stops when it detects plant stems and opens its manipulator step by step so that the Kinect v. 1 camera can capture deep RGB11- Red, Green, Blue. colored images from the various sections. After calculating the bulk volume of the plant, it starts spraying. The results showed that the robot is capable of calculating the volume efficiently. The average operation time of the robot detecting plant, opening manipulator, imaging, estimating volume, and spraying liquid according to the estimated volume is 54 s for a plant of 1.7 m height. The detection error of the robot is estimated to be 19%, which is negligible. Assessment of liquid-sensitive papers located around the plants shows that the robot’s liquid-spraying quality is not reduced at high levels of the plant. The robot can detect the plant’s height and preserve its desired spraying quality in its total height.